CA2204359A1 - Method for making recombinant yeast artificial chromosomes - Google Patents

Abstract

The present invention provides methods for construction of recombinant Yeast Articifial Chromosomes ("YAC") by homologous recombination between YACs during meiosis. In particular, conditions are provided for the step of mating haploid cells and for the step of spore analysis that increase the frequency of spores containing the desired recombinant YAC. The methods find particular use in constructing recombinant YACs between YACs that are incompatible when copropagated in a diploid and/or that share homology regions of less than about 50 kilobases. Linking YACs, methods of their construction, and methods of their use are provided that allow facile construction of a YAC containing two or more discontinuous regions of DNA.

Description

1.

METHOD FOR MAKING RECOMBINANT YEAST ART~lCIAL
CHROMOSOMES

ACKNOWLED~ ~TS
This invention was su~polted in part by grants from ATP grant number NIST CRADA 70NANB3H1366 and from NIH SBIR grant .. I-er NL9ID
AI32285-03. The U.S. Gove-. -.. -~1 may have rights in this invention.
INTRODUCTION
Technirq1 Field This i,l~.,ntion relates to m-othr~c of construction of Yeast Artificial Chromosom~c ("YACS") by homol~ous l~..lb;nAI;~n Bacl~ul~d Yeast Artificial Chromosome (nYAC") cloning vectors are capable of pro~qg?ting hrge (50 to more than 1000 ki1nbqces) cloned inserts (U.S. Patent 4,889,806) of ~.nf~f.~ '. DNA. To date, an upper limit for insert size has not been ;n~l YAC clone 1;1-.A-;eS have been used to id~.~lify, map, and propagate 25 large r,A~ nl!i of ..,---....qliqn g~nnrnic DNA. YAC cloning is espe~iqlly useful for iCon ~ing intact genes, particularly large genes having exons disllil.u~ed over seve~al tens of kil~qces or more, and genes having distal reguhtory c4....1.nl~ located tens of l~lobases or more u~ ~ or do~. "~1,~" from the exonic s~uences YAC
cloning is particuhrly advantageous for isolating large complex gene loci, such as 30 ulll~A~.AI-ged i~ gl~ulin gene loci. YAC cloning is also well-suited for mal~ng vectors for p~,rOl",il~g ~led homologous recombination in ~z~ liqn cells, since YACs allow the cloning of large contiguous s~l,~n~es useful as recombinog~.-;c homology regions in homologous ~;eLil~g vectors. Mol~"
- YACs afford a system for doing targeted homologous recombination in a yeast host 35 WO 96/14436 PCI~/US95/14966 2.
cell to create novel, large l~ sgelles (e.g., large minigPnps~ tandem gene arrays, etc.) in YAC constructs which could then be transferred to .. z.. qli~n host cells.
Unlollundl~ly, manipulation of large polr~r~4t;~les is problematic.
Large polynucleotides are susceptible to breakage by chP~fing forces and form highly viscous solutions even at relatively dilute conr~ ~ l;ons, making in vitro 5 m~nirllqti~n eY~-ee~l;n~ly ~lifflrlllt For these reasons, and others, it is ciesir~hle to reduce the amount of in vitro manipulation that YAC clones and other large DNA
f~gmPntc are subjected to in the ~.~cess of constructing large ll~s~ene constructs or homologous ,~...k;..~'ion constlu~ls Cloning of large fr~nP.ntc of genomic DNA in YACs has become a 10 gene~al approach to study the physical org~ on of complex g~l~o,l,es (Burke, et al. (1987) Sci~-nre 236:806:812). The a~ .ge insert size of cu~ ly available YAC libr~riP.s varies l~wee~ 250 kb (Burke and Olson (1991) ~-P-th~s Enymol.
194: 251-270) and more ~ltly 850 kb (R~P.llqnP.-ChqntP~1~)t et al. (1992) Cell 70:1059-1068). Some genes (CFTR, BCL2, DMD), or their regulatory- el .~ , 15 extend over several hundreds of ldlobases, and therefore, are not always going to be present within a single YAC. Moreover, the ~l~Pt~ analysis of seveIal YAC
;CS has revealed a high ~J-x~ ge (up to 50%) of chimPri~ clones (Green and Olson (1990) Nature 250:94-98), implying that even positive clones Cal1~ g largeinserts, might be of limited usage. 20 ~ Pntly, meiotic homologous l~c~ ..-hi~ in yeast has been used to l~conshuct larger yeast artificial cLvl,,oso,lles starting from a diploid yeast Cal1,~1ng two o~e-l~r~ping YACs. Using this al)p~ach, Green and Olson ~bid.) were able to build l~;co...b;n~n~ YACs up to 790 kb, lhe.~ conl~ g about half the DNA ;CQ1~ from the cystic fibrosis gene region on human CL1~J~I-OSO~-P~ 7. 25 Since the parental YACs were Cd11,~ g ifl~.nti~ql pairs of sel~tqhle ...~ (I~Pl and URA3), the recombinqnt clones were i~ ;r~d by using scl~enil~g by physical assays (e.g. PCR-based ~se~.nPnt of previously defined se~uen~d-tagged sites (nSTS") cont~.nt, and size measurements by pulsed-field gel d~1l0~hol~is). Priorto recombining, YACs were analy~d for common regiohs, i.e. o~e.ld~s, and 30 diploids co~ inin~ a pair of YACs with appl~l at~ly o~ la~p,ng (i.e. sharing long regions of homology ori~nte~l in the s. me direction with respect to the YAC arms) WO 96/14436 PCTtUS95114966 3.
regions for meiotic recQmhinqtir~n were constructed. One recombinqnt YAC, pCF-1/7-R, of about 600 kb was constructed by recomhin-qti~n of two YACs l~olledly having an overhp of 40-50 kb of homology by scl~ng 76 spores. Cellini et al.
(Nuceleic Acids Research (1991) 997-1000) reported a YAC vector (PYACR) with LEU2 as an du~ hic marker on the right arm, but with no ...~ on the left 5 arm. The genetic s~l~li~n l~ lcd by Cellini et al. for one of the recomhinqnt YACs is ~l~s~tly of li ited usage due to the hck of avaihble YAC l;l..,..;fs constructed in a leu2~ host. None of the above m~th~s lq)oltcd sucoes.~ful ~cco..~ n of YACs sharing less than about 40-50 kb of homology nor of pairs of YACs that were obse.~ed to be mitotically inco.. ~ during ,n~ t~ to 10 o~agate them in the same diploid.
Sears et al. (1992 PNAS USA 89:5296-5300) d~ ~loped a YAC
based recombi~ ion system to study factors co~ ;l,uling to the fidelity of meiotic chromosom~ ;cs;on They were able to show that the rehtionship bc~n physical ~iict~ncc and recomb;~ ;. n r~u~n~ within the human DNA s~g, ~-.l 15 insert was co...~ '~ that of en~o~n(~l)s yeast ch,u...~s3c,...~1 DNA (2.0 to 7.7kb/cM). They used .l,e;o~ic recombin~tion bciwecn YACs pl~S ~ 50-360 kb of overlap. Only ~iirloi-l~ co.~ini-~g pairs of YACS that were mitotically co...p~ihlP
were used.
Den Dunnen et a/. (1992 ~um~n Moleculqr C}PnP-ti~s 1:19-28) also 20 took advantage of hG.llGlog~Jus ~G~Illbination to l~COll~uCt the majority of thehuman DMD gene in a single ,~...hi,.~.,l YAC of 2.3 Mb. They used meiotic recombination ~l~ YACs ~ ,s~ -~1;~ 150~60 kb of overlap and that were mitoti~lly cn~ t)lp in the .liploid However, Den Dunnen obsel~ed that some pairs of YACs did not lead to i~olqtion of desired recombinants because mitotic 25 pror~tion of the YAC pair in the ~iirloi~, prior to mP;osi~, resulted in the loss and /or gross ~lal~g~,l,ent (observed as a change in size) of one or both YACs;
this behavior was l~f~l~d to as diploid il~co...p~;hility. Reco...b;~ ion was not obt~ined with YACs that were found to be mitoti-~lly inc~ lpl~lihlP when a~ s were made to co-prop~te them in a ~lirlo~ 30 Cul,~ the intro~Uction of mllltiple and ~licrP~e genes into a target cell or animal ~P~ es se~ e ~,ansr~iLio~ ,g, ~lion steps for each gene 4.
typically followed by tedious s.;l~n-ng and breeding procedures to derive a l,~sge,l~ic cell or animal having each of the sep~.,.le genes. Coinjectirln of the nm1tir1e and ~i~rp~e genes is problematic in part beclq~use a resnlting ~ ee~
aIray (q~ ...h~g one would achieve gen~Ptil qlly linked co;~ n) would be unpredictable and uncol~ llable, such that the stoi~hiometry of the COIII~G~ 5 genes of the t~ -see~-P array could not be controlled, and structural analysis of the trqn~gçnP. array would be very ~iffi/'111t if not i,..~ossible to achieve.
Thus, there exists a need in the art for an effi~iPnt and versatile mPth~ of pl~a.i"g Large se~ of DNA in YAC clones with a .-.;,~ of time, m-q-nip~11qti-~n, and cloning p~cedul~,s. In particular, it would be highly lO
advantageous if it were possible to obtain cloned .. -.. ~liqn genomic fr.qgmPntfrom a YAC libIary by recombiL~ on of YACs without ad-liti~nq1 cloning or manipulation (e.g., 1ie,qti-)n of the sellupn~es to each other), with ...h.;...~1 s~;l~nil~g to find al~pr~ Llely ~I~,~ldppi~g YACs, and willloul the need to pre-screen for, or be reliant upon, diploid co...p-~;hi1ity of YAC p irs. Such a recomhinqnt process 15 would be useful to construct larger yeast artificial chru--loso - es from sm. ller o~ g ones, to e1;~ e the rhimPric parts of some YACs, to ~on~Lluct a clone c~n1qil-il-g a large genomic region of interest, and as a means to construct large DNAs of a desired design, for example, one that would combine genomic regions nqtl1rq11y s,pa."t~ by un~lp~irq~blp or unt1onqh1~P regions or one that would 20 create clusters of genes that are not normq11y grouped.

SUMMARY OF THEINVENTION
It is an object of the present invention to provide mpth~s for constmcting large DNA constructs as Yeast Artificial Chromosomes by homologous 25 recomb;n~ n b~wee n YACs(~f~.lcd to as parental or input YACs) during meiosis in ye. st. In general, homologous recomhinqfion l~ween a pair of YACs that share a region of homology is ob~il~ed from a process that inrllldes a mating step, a sporulation step and an identifi~-q-tion step. In brief, the process in~.lnr~es the steps of (l) f~st mating yeast cells of o~osile mating types, where each type 30 cc.-~-q-;~-c one YAC of the YAC pair to be recomhinP~, to form a diploid that contains both input YACs, then (2) sporulating the diploid, i.e. inducing the diploid S.
to undergo meiotic division and haploid spore formation, usually by ~llU~
starvation, during which recomhin-q-tion l~weel~ the YAC pair can occur, and (3)s.lbs~l~- n1ly idcnlilyii~g (and i~olqtin~) those hqpk.i-l spores bearing a desired recombin~qnt YAC. ~lthnllgh the invention can be applied to recoml.;..-l;. n n any YAC pair sharing relatively large regions of homology or ~ a 5 YAC pair displaying mitotic stability when co-proFagqtP~ in a diploid, the mpth~ls of the invention find pa~ticular use when applied to YACs with hom~ gy regions of less than about SO kb, ~,~f~ably less t_an about 40 kb, and/or, most preferably,that disphy mitotic i~4...~ ;hility when a~ s are made to co-prop~;qtç them in a dirloi~l~ Accol-lh~gly, it is an object of the invention to provide mPth~S that 10 inc,~ase the frequency of haploid yeast spores cn ~1 ..n;n~ the desired recombinqnt YAC after the spomlation step, and to provide ~ lhn~l$ that select ~ f~nLially for desired h~ d cells CC~t~ ;~ the desired recomhinqntYAC from the spore populqti~ n to thus ~n~ dse the rl~u~ l~;y of desired cells in the population of cells to be analyzed for the ~ ,sence of the desired recombinqntYAC. Accolding to the 15 invention, mP.th~l~, which include steps for consL,u-1ion and mqnir~llqtit)n of dirloids prior to sporulation and steps for subs~l~e~l genetic sçlection of YAC-co.~ ;..;. 2 spores, which steps can be used s~q.~.ly or, most ~l~f~ably, in combination to best acl~icve the rcl~oil~ objects.
Accc.,ding to the invention .-.~.~s for diploid construction are 20 provided such that during the mating step a diploid yeast cell, cc..~ iug a pair of parent (or input) YACs that share a homology region ~le to ena~le homologous n l~lwecl~ them during meiosis, will ...~ both input YACs willluu ~ulange~ t upon e-he. ;l~g the sporulation step. Ihis is ~co...pli.ch~1 by ...h~;...;,;ng or p,~,c~Ling diploid growth, i.e. mitotic division (doubling) of the 25 diploid cell, ;"".,~ 1y after mating during the mating step. Diploid growth during mating is limited to about 8 or fewer diploid doublings or most preferably c;vell~ed. When mass ...~t;ng~, i.e. mixing of cultures of the parental haploid yeast cells, are ~lrol",cd, mitotic growth is most preferably pl~e,lled or is limited by l;...;l;ng the mating period to a period of time equal to or less than 8 diploid- 30 doubling times. ~f~dbly mating is pe.Ço"l,ed in the absence of selection for ...5.L...:i located on the YAC anns. Recombin~q-tion (meiotic recoml);.~;c.") is 6.
in~ oe~1 by sporulating the diploid or ~lip1nitl~ The res~lting cell ~ W~, co..~ .;ng spores, ~lirl~ , and lmmqtPA pa,~,~led hqploi(ls, is then enrirh~ for(haploid) spores. The spores can be sc~l~ed by the usual methn l~ in the art fort_e desired recomhinqnt YAC, or preferably, a mPthod of genetic S-P1Pcti~n acco.~ling to the ill~enli~n, as pl~s~ d below, is followed in order to select for and S
enh~nce i(l~Pntifirqtion of spores co.~'~il-ing the desired YAC.
Meth~l~ are provided for the spore analysis step that fr~i1itqtP, d~ - and i~ n;rl~ation of spores having the desired recomhinqnt YAC by p~.~ol~ ng a genetic s~ ;n n for yeast spores co.~l~;nh-g the desired recomhinqnt YAC, whi1e optionally cuun~r s~kc! ;n~ (i.e. sPlPcting against) for parental YACs 10 or the u~desil~d by-product ~o...1);,-~n~ YAC. Acco~ding to the invention, genetic selPcti-n is d~;gl-~l by providing t_e input YACs with se,lPct<q~b1e ...~.k~ such that YAC recomh;~';n n results in the desired recombinant YAC pre lirtql 1y having a certain sekPct-qh1-P marker, or preferably ..-~ " that allow app1irqtinn of a genetic sel~Pctinn for the marker or ~ L~ ~ to allow growth of spores CQ-~1; in;i~g the desired 15 recomhinqnt YAC and optinnq11y to select against the marker or ...~.k...., present on the undesi~ l~hlct recomhin-q-nt YAC, or fur~er optionally select against cells with parental YACs. After diploid construction, ~ ~ably by a mP.thod of diploid construction accol.lh~g to the present ill~,e.llion, and sporuhtion, spores are enrirhed with respect to diploids and u~ ed parent h ~r1Oi-ls by mPthn(ls known in the art, 20 and the ~ ...;I h~ spore population is sul,jected to genetic sflF~1;0l- in media that selects forthe desired leco.-~l-;n~ YAC and, optionally butpreferably, counter-selects for the unde.,i~ by-product ~...h;n~.~1 YAC and further ~ptinnql1y but most ~ f~.ably cuu~lt,r selects for parental YACs, (e.g., spores with unl~o~llbined parental YACs residual parental diploids). As a con~u~ .-re of the mpthnd of the 25 tiol~, the res~lting ~ ing cell population has an increased rl~uel'cy of har1ni-1 cells co~ ;ning the desired YAC, and accoldh~gly, the s~se~lue,ll SCl~l~g and analysis for cells ha~ g the desired recomhinqnt YAC is greatly simplified and ~l~l;t~
Although either the step of ,..;~ ;. n of diploid mitotic growth or 30 the step of genetic sPl~tinn~ pl~f~ably sPl~ctinn for the desired recombinant YAC
and against undesil~ d YACs, can be pe.rollllcd without the other as a step in a 7.
known YAC recomhinqtion protocol, in the most pl~f~lcd embo~lim~nt of the invention both steps are p~rulll,ed.
The invention finds particular use in constructing large DNA
constructs of ~ g~nfS and large hom-llogous l~g~ g constructs ~)~nn;,~ at least one comrl~~e l.i.n.~r.. ;~ 1 complex, suit-q-hl~- for tl~sr~,. into ~.~D.. ~liqn cells, S
such as embryonic stem ("ES") cells, typically for construction of a !.,.n~genirammal. The m~th~s of the present i.l~cnlion also find particuhr use in constructing minigenP.s that co-..l~ e rl~lsters of genes that are not n-)rmqlly found clu~t.,.~,d in a gc--o.nc or that are located in dirr~ t genomps.
Furthrrml)re, the u~ nlioll~ particuhr linl~ng YACs and their 10 m.eth ~A~ of use, finds use not only in adjoining large regions of discc~ o~ls DNA, but also in G~g;n~ g hrge DNA constlucts by adjoin-~g DNA constructs such that nn~ hlP. regions in the starting mqtrriql.~ are eyr1nded from the final recombin-q-nt product.
Finally, novel m~.tho-ls of making the linl~ng YACs of the invention 15 are provided that .n;~ .e steps involving DNA isolation and host cell srJ....~ n BRIEF DESCRIPIION OF THL DRAWINGS
Figure 1 is a sc-he~ ;c depi~;ng re~mb;n~ n bclween YAC pairs 20 Jl.3LYS and YNN-~y-HlS to produce Jl.3 ye2. The YAC vector arms are e.s~-n;~lly as de~.;l~ed in "YAC Lil,l,ui~s, a users guide", Nelson and 13~v~nstGin, eds. All the YACs carry trlr....,-;r e1r-~ nl~ (~en-)tP~ by heavy arrowheads) at the ends of the YAC. The URA3, TRPl, HIS3 and LYS2 yeast sPlP~t~hle marker c~cc. l~l,s are d~P-I-otffl as striped, open, dark, and dark boxes l~e.;tively. The ARS 25 c~cs~P.ttç n~j~~Pnt to the TRPl c~ccette is depicted as a grey box. The open circle m~rkP"l "cen" dPn-Xes the yeast C~ G~ while the striped and filled boxes m~rkP11 "ampr" and "kan"' denote the bactP.~ mpir.illin ~ ;."r~ gene and the yc~ re~ict~n~e gene l~tively. R~Ppl~~emP.nt of the URA3 marker on Jl.3 with the LYS2 marker by targeted l4co.. h;~ n of the Jl.3LYS c~csettP~ is 30 dP~icP~ on Parent YAC One. I2PP1~~~n,~n~ of the TRPl marker with the HIS3 marker by ~eled lGco...hin~ n of the YNN~yHIS c~cc~llr is depicted on Parent WO 96/14436 PCT/US95/14g66 8.
YAC Two. SpeI and NotI lesl~ .1 ion sites are shown at the ends of the YAC inserts and are denoted "S" and "N" ~ e;live1y. A fine structure l~s1~ n map of J1.3 is given in Choi et al., Nat. Genet. 4: 117 (1993). The f~ed oval in the S y1-C y1 region fr.qgmPnt d~n~tes the rat heavy chain 3' P~-h~n-~,r. Par~al restriction maps of each input YAC as well 5 as the res~ lting recomhinqnt YAC products are p,~se~led Figure 2 is a s~ depi~;ng recoml~ n ~Iweel~ YAC pairs J1.3LYS and NS10-B14. Partial l~,s~ n maps of each input YAC as well as the res ~lting recombinant YAC products are pl~ ~lr~. NS10-B14 is a linking YAC.
Figure 3 is a s~ ic rlep;~t;.~g re~oml.;ft~-~;f n l~c;lw~l~ YAC pairs 10 J1.3-B14 and P1-570-2-1. Partial .~sl~ maps of each input YAC as well as the res ~lting ~---1);-~--1 YAC products are yl~s~ ~1ed. RecomhinqntYAC J1.3-570 is an ~.;....pl~. of a desired ~comhin-q-ntYAC res~1ting from a multi-step process of the i~ ioll wh~ n a linking YAC, (see NS10-B14, Figure 2) is used for producti-n of an ;.,t~ e YAC (J1.3B14). 15 Figure 4 is a sCl~- .nAIi~ depi~ting recoml)in~;oll b~ YAC pairs C13-X15 and J1.3 ye2. Partial restrirtion maps of each input YAC as well as the res ~ltin~ recombinant YAC products are yl~,s. ..1PA
Figure S is a scl~ ;r de~i~ting recoml)inali~n ~lween YACs C13-X15 and J1.3B14. Partial restri~tion maps of each input YAC and the res~lting 20 YAC plOdll~lS are ples~
Figure 6 is a s ~ d~p~ ~ a pn~locol for recombination of YACs with short regions of homology. The small and large rounded-corner -g1Ps denote h~rl~id and diploid yeast l.,specli~ely. YAC arms are dello~ed as clear boxes, with I ~ ;angl~s at ends denot;ng ~lo,~ ,s. The insert homology regions 25 are dep: ~ed as grey boxes.
Figure 7 is a map of the human immlmoglobulin heavy chain variable region. The p1 ~Pment of the YACs with respect to the vafiable gene scale is a~l~ e.
Figure 8 depicts par~al re~tficti~n maps of p1~Qmi-ls 24.13 and 10.33. 30 pl~mids are depicted after li-.~ n with XhoI.

9.
Figure 9 depicts construction of a linking YAC insert by LR-PCR.
The four p. ;.. ~ for LR-PCR of en-lr1~nP inserts (S3CX, S2CRl, S2' AXl, and SlAR) are depirted as shaded ~ nglPs. S3CX and SlAR contain NotI sites.
S2CRl and S2' AXl contain i-SCEI sites. Digestion of LR-PCR prûducts with i-SCI, 1igqti~n, and ~ C~ n with NotI yields th~e p,~o...;nAn~ dimer rn-~lPCuhps~ 5 the desired h~ r (den~tp~d (a)), the 10.33 homodimer (denotPA (b)), andthe JlXK.31 homo limPr (denoted (c)). Product (a) is i~olqtA from (b) and (c) by size fractionation by p~q)a~alive pulsed field gel el~tlul~h~ ,is. Pnmer sP~u~Pnces are provided in PY- , le 11.
DES~Kl~llON OF SPEC~lC EMBODIMENTS 10 Unless defined otherwise, all P~chni~q1 and sc;~ ;rlc teIms used herein have the same ...~I.ing as commonly understood by one of o,di~ skill in the aIt to which this ill~ n belo~. ~hht llgh any mP.thn-l~ and mqtPriql~ similar or equivalent to those de~~ ed herein can be used in the pT~ tice or testing of the present ill~/Gll~ the pl~f~l~d mPth~ and mqtPriql~ are desçrihed. For ~ul~oses 15 of the present illvGllLion, the following terms are ~PfinP~d below.
The term "coll~spollds to" is used herein to mean that a polyn-lclt~ e s~u~nre is homologous (i.e., is irlPntirql, not strictly evoluLiona,ily related) to all or a portion of a rGfelGllce polr~lr1~otil1e setluence, or that a polypeptide s~lu~ e is ide--l;- ~1 to a ~f~n~nce polypeptide se~luPn~e~ In 20 contrs~ l-, the term "compl -- .~ to" is used herein to mean that the comple...~ s~l~t ~-~e is homologous to all or a portion of a l~f~nce poly..--cl~l;t1e se~u~Pn~e. For illu,,l~t1;0n, the nUCl~Poti~lp~ sequ~Pn~e "TATAC"
co"~sl,onds to a l~fe~nce se~U~Pn~ e "TATAC" and is comI l~Prn~Pntqry to a l~,f~ ce s~u~ n~e "GTATA." 25 The tenns "~,uls~ lly col,~,~,~nds to", "sul,~ t.~ lly homologous", or "substq-ntiq-l idGlllily" as used herein ~lPnotes a chqr~ctP~i~tic of a nucleic acid snllJ~Pnce, wL~.~.n a nucleic acid se~lu~-r-e has at least 70 ~ ;Glll s~~u~ e identity as coll~ cd to a lcr~ lce sequenre, typically at least 85 ~elll sequen~e identity, and plGÇ~ldbly at least 95 p~ .;Gnl seque~ce identity as COll~ d to a lGr~ Gnce 30 sequP-nre. The p~enlage of sequ~P.n~e identity is cql-~ulqt-P~l eY~ ing small deleti~n~ or ~1-1iti- n.~; which total less than 25 pf.~;Gllt of the ,,~f~.~nce s~P~u~.nr~.

WO 96/14436 rCT/US95/14966 10.
The ,c;r~ce sP~uPn~e may be a subset of a larger sequence, such as a portion of a gene or fl~nking se~u~ e, or a l~pctili~e portion of a chromosome. However, the l~f~nce se.luell~ e is at least 18 m-cl~oti-lP,s long, typically at least 30 nucleotides long, and pl~f~ably at least 50 to 100 nucleotides long. "S~ 11y co.. ~ .. k-~ as used herein refers to a s~lu~n~e that is co.u~1r .. ~ . y to a 5 se~u~ e that ~ubs~ lly co"~,s~nds to a ~cr~ "ce se~upn~e.
SpP~ific hybridization is defined herein as the formation of hybrids bc~ a ~_Ling !.,..~.~g.-l-~ s~l.,e~ e (e.g., a poly. ~~1FA~ e of the i~ ion which may include s~ilulions, de~ l;nl~, and/or ;~ltliti-~n~) and a sper-ific target DNA s~lu- n~c (e.g., a human APP gene s~l.,ence or human ;.. J.-oglobulin gene 10 s~l~en~e), W}l~ .n a labeled ~e~ g ~ladsge.~e se~u~pnr-e ~lvf~.vnli~lly hybri~li7Ps to the target such that, for v~ 1r, a single band collv;,l~n-1;ng to a restrir-tion frq~nPnt of a gene can be id~ d on a Southern blot of DNA plvpalvd from cells using said labeled ~vt ng l~ C~n~- sequen~e as a probe. It is evident that optimal L~ n c~-n~ n.c will vary ~lepel-~ -g upon the s~uen- e col.l~silion 15 and length(s) of the ~v~lg ~sge.le(s) and endogv.l~us target(s), and the e~ ontq1 method s~Fcted by the pra ti~llv.. Various gui~lP1in~s may be used to select app~pllate hybri~ n~ ;ol-c (see, Maniatis et al., Molecular Cloning: A Laboratory Manual (1989), 2nd Ed., Cold Spring Harbor, N.Y. and Berger and Rimmel~ Methods in Enymology, Volume 152, Guide to ~o1Pcu1qr 20 Ck~nin~ Te~hni~lues (198n, Ac~çmi~ Press, Inc., San Diego, CA., which are ~co~ dted herein by l~fe.~nce).
The terrn "rq~nr~q11y oc.;~ " as used herein as applied to an object refers to the fact that an object can be found in nature For ~...1,1e, a polypeptide or polym~r1~oti~ sequ~-nce that is present in an o.~nic .. (inr~ 1ing vimses) that 25 can be i~s1qted from a source in nature and which has not been intPnti~nq11y mo iifi~ by man in the laboldtol ~ is nqh1r.q11y~ g. As used herein, laboldtol~ strains of rodents which may have been selectively bred acco~mg to c1q~ir-q-1 g~qn~tics are considered nqh1rq11y-oc~ qnimql~.
The term "c~ P" as used herein refers to a gene s~u~.~ce that is 30 evo1ution-q-ri1y and fi~n~;l;o ~lly related bcL~ ~ies. For eYq-mp1P but not 1i...ils~inn, in the human genome, the human i . ~ nog1nbulin heavy chain gene 11.
locus is the coglldLe gene to the mouse ;.. J--r,globulin heavy chain gene locus, since the s~uç~ es and structures of these two genes in~lirqte th. t they are highly homologous . nd both genes encode a protein which fimrtir)n~ to bind qntigPn~
lly.
As used herein, the term "xcnoge~c" is defined in relation to a 5 i~ n .. s.. ~liqn host cell or n~-nh-lmqn animal . nd me-qns that an amino acid s~l"~.n~e or polrllrhPoti~le s~P~uPnre is not encodP~l by or present in, l~ ely,the nq~nr.qlly~;.. ;ng g~o.. e of the recipient .. ~.. ~liqn host cell or n-nhnmqn anim. l. Xcnoge,.~ic DNA se~ -r~5 are foreign DNA s~lvenr~s; for eY~
human APP genes or ;.--....~--~l~ulin genes are YPnogPnir with respect to murine 10 ES cells; also, for ill.-,~ ,-, a human cystic fibrosis-qcsor~ d CF~ allele is ~nog~,.~ic with respect to a human cell line that is homozygous for wild-type (normal) C~l~ alleles. Thus, a cloned murine nucleic acid s~uenre that has been ~---ul~1~ (e.g., by site di,~ted mnt-q.~PnPs;~) is yf~ ir, with respect to the murine gen.. r. from which the s~u~--re was ori~nqlly derived, if the .. ~ ed sequenre 15 does not nqtnrq.1ly occur in the murine ge-~ r, As used herein, a "heterologous gene" or "heterologous polrnrl~4ti-le S~lu- ~re" iS definçd in l~,L~n to the ll~sge ~ic nonhllmqn ol~;anis producing such a gene product. A heterologous polypeptide, also referred to as aA~nGg~,~e;c polypepti~e, is d~PfinP~ as a polypeptide having an amino acid se~u~P.nre 20 or an encoding DNA se~ .n~e coll~ e to that of a cog~le gene found in an o.~ni~... not con.-:-';ng of the l~ e.~ir ..onl.~ - animal. Thus, a ,~ ee.ic mouse l~l~ling a human APP gene can be desc. il ed as h ll~~ g a heterologous APP gene. A l~,.n~g~1~ir mouse hall Gl~g a human ;.. .~ globulin gene can be deP,e~-hed as ha.lo~ , a heterologous ;.. ~ f)globlllin gene. A !.~ gf.l-l~. 25 conl~;--;--g various gene s~ e~r~~ a heterologous protein seqUpnre may be readily ~ 1;r~pd~ e.g. by hybri-1i7qtir,n or DNA sequPnring~ as being from a species of O,~p~ ... other than the l.~ ir. animal. For ~A~ull~)le, eAl)~es~ion of human APP amino acid s~u~ nres may be detect~P~ in the ~ sgenic nl-nhl-mqn qnimql~ of the hl~ ion with antibodies specifir~ for human APP epilo~s enrodP~d 30 by human APP gene se~ pn~. A cognate h~,t~,lvlogous gene refers to a COll~ spol~ding gene from al uth~r species; thus, if murine APP is the l~,fe.~ ce, 12.
human APP is a cog~ e heterologous gene (as is porcine, ovine, or rat APP, alongwith APP genes from other species).
As used herein, the term "~eling construct" refers to a po1yn~lrleoti~e which CG...~ S: (1) at least one hnmology region having a s~ n~-e that is S.lbS~A.~ 11Y i~ 1;r~l to or sulJ~lA.~ 11y C0"'1)1~ A~Y to a s~uence present S
in a host cell en-lGg~--o!~ gene locus, and (2) a ~t~ g region which becoll,es ;i~lr.~ into a host cell end~g~(,us gene locus by homologous recoml)in~;o ~t~cc~ a I~E,ling const~uct hnmolngy region and said e---dog~ s gene locus se~lu ~e. If the ~_t;,lg construct is a "hit-and-lun" or "in-and-out" type const~uct (V~l~n~ills and Smithies (1991) Mol. Cell. Biol. 11: 1402; Donehower et 10 al. (1992) Na~ure 356: 215; (1991) J. NIN Res. 3: 59; Hasty et al. (1991) Na¢ure350; 243, which are i~co~ led herein by l~f~ce), the ~ c~ g region is only I~A~C ~ Y i~lcol~l~ted into the e-~Gg~ ~J-~s gene locus and is ç1;~ led from thehost genome by s~ n A ~ling region may CO1~1~1iSe, a se~uence that is ,S1A~ 1Y hnmo10gous to an ~lldog~llous gene s~u- n~ e and/or may col .i~ a 15 n~nho..~o1ngous s~f~ enre~ such as a s~ t~1P marker (e.g., neo, tk, gpt). The term "~etillg consl~u~t" does not n~es~.;ly in-iir~te that the polynucleotide co---l--ises a gene which bec-,lllcs ;nt~A~ffl into the host genome, nor does itnecesC-. ;1y indicate that the polynucleotide COlllpli~S a complete stmctu~al gene s~lu~nce. As used in the art, the term ''l'Alge~ p construct" is ~lWI~llloUs with the 20 term "~geli.~g l.,.n~genf." as used herein.
The terms "homology region" and "homology clamp" as used herein, when l~f~ri~g to a ~E,_ling constIuct and an ehdGg~l~ us gene se~enre, refer to a s~...f.~1 (i.e., a portion) of a ~_~lg constluct having a seq~lenre that sul sl-Anl;~11y colles~onds to, or is ~"b~1A-~ 11Y comrl~m~nt-A1y to, a P1~1~ -.. -;--fl1 en-log~.n-us 25 gene sf~l.,~ r~, which can include se~u~ nreS fl~nting said gene. When l~ fell~ to homology regions shared bGl~oen two YAC constlucts, "homology region" refers to a se~ en~ of one YAC construct that sub~ lly collG~onds to, or is ~ub~ lly co...l~ tont~ry to, a region on the second YAC construct. A homology region is generally at least about 100 nurleoti~es long, preferably at least about 250 to 500 30 nucleolides long, more preferably at least about 1000 n~ eoti~les long, or pl~,f~lably longer. ~lthough there is no ~l~m~ eol~Lical ~;n;~ u length for a L~.
homology clamp to ~,.erl;~tP~ homologous recombi~ , it is believed that homologous recnml-in~;. l- effl~ipnry geneIally increases with the length of thehomology clamp. ,~imilqrly, the ~ ;n~ çffi~iP.nr.y increases with the degree of sequpnre homology 1~YCeI1 a ~t;ng construct homo1ogy region and the c~dog~ ,us target s~ nr~, with optimql rec~lm1);,ul;~-l- effi(ipnry o~u~ g when 5 a homology clamp is ;CQgPn;r. with the enrlQgPnnlls target seql~Pnr~. The terms "hnm~l~y clamp" and "homology region" are h~ Ang~ as used herein, and the q~ e t~ .inology is offered for clarity, in view of the inr~n~iQt~Pnt usage of similar terms in the art. A hnmology clamp does not ~e~.~1y co.-note formqtinn of a base-paired hybrid structure with an endogenous s_quPnc~P. 10 E~doge~l~)us gene s~ ees that s-~b;,~;..l;~lly co~ ,~nd to, or are s~bs~ 11y c~ to, a l.~n.~.n~. homology region are l~fell~d to herein as 'ICl~)SSv.target sequPnre~" or ~n~Gg~ n.JI~s target se~ r~~. "
As used herein, the term "minigenP." or "miniloclls" refers to a heterologous gene construct wLc.~ one or more ~nnf.ssc..l;~l seg~ nl~ of a gene 1 are deleted with respect to the nq~llrally-oc~ p gene. Typically, deleted se, ...-~ are intronic .~ ue~ces of at least about lO0 bq~ to several kilobases,and may span up to several tens of l~lobases or more. Isolation and manipulationof large (i.e., greater than about S0 ki1~q~P,~) I~ g constructs is rl~uenlly ~iifflr~llt and may reduce the er~ e ~c~ of l-~lsr~ g the l~eling constluct into a 20 host cell. Thus, it is L~u~ e~ f. to reduce the size of a l~;eLillg construct by ~lf1el;-~g one or more ~ nfsse ~ 1 pollions of the gene. Typically, intronic s~lueMr~~ that do not cnco~ )ass es.~ 1 regulatory e1~ may be ~lP1eted. For e, a human ;.. -.. ~1Obl.1in heavy chain minig~nP may co~ ),ise a ~-P1Pti~n of an intronic s~m~-nt l~l.ve~.l the J gene seL...~ and the ~ COnSt~lllL region exons 2S
of the human heavy chain ;,....~.-~]~!~1in gene locus while l~,~;.;n;~g illll)olt~
re~lqtory e1~-mP-ntc in that region. In an additional eYqmp1e of a minigene, a human APP minigene can comrri.ce the spliced exons l to lO of human APP
(obtained from a CDNA, thus having introns removed) joined to exons ll to 18 ob~ined from APP ~nnmir DNA. rlc~lut;lllly~ if co,l~ient restrirti-n sites bound 30 a non~se-~ 1 intronic ~e~uenre of a cloned gene se~uenre, a deletion of the intronic se~ onre may be produced by~ ligesting the cloned DNA with the - 14.
a~ e restrietion enzymes, (2) sep~ g the restriction frqgmP.ntc (e.g., by electrophoresis), (3) icolqting the restri~tion fragmP.nts ellco...l ~ccing the esc-pnl;~1 exons and regulatory elPmpntc~ and (4) 1ig,qtin~ the isolated restri~tion fr.qgmP.ntc to form a ~..h~;g~n~ wl,~.~.n the exons are in the same linear order as is present in the germ1inP copy of the nah~r~q.11y-oc~;lll ~ ;i~g gene. ~1tPrnqtç methods for producing a S
miniEenP. will be appalG,lt to those of skill in the art (e.g., ligation of partial genomic clones which encol~ass esc~ 1 exons but which lack portions of intronic se~uen~e). Most typically, the gene se.gn.e.~1~ comrricing a mini~,PnP will be ~ ged in the same line. r order as is present in the geImline gene, ho~ er, thiswill not always be the case. Some desired regulatory elem~Pntc (e.g., Pnhqnçers~ 10 .ci1P.n~erc) may be relatively position-;n~e ~c;l;~e, so that the regulatory elPm-Pnt will run~ ll coll~ctly even if po~ lirr~ lly in a minigene than in the co~ i,~ndil~g gP.rm1ine gene. For eYqmp1e, an e ~h~,re~ may be located at a dirr~ rlictqnce from a promoter, in a dirr~ l o. ;~ , and/or in a dirr~l~nl linear order. For eYqmp1P, an enhqnr~r that is located 3' to a promoter in germline 15 configuration might be located 5' to the promoter in a minigPnP. Similarly, somegenes may have exons which are qll .n~;vely spliced at the RNA level, and thus aminigene may have fewer exons and/or exons in a dirr~ t linear order than the co~ ollding germline gene and still encode a functional gene product. A cDNA
~n~ g a gene product may also be used to constluct a minigp~ne. However, since 20 it is ~enP,~11y desirable that the heterologous minig-Pne be eA~l~,ssed .cimil~rly to the cognqtç nq,tnrally-oc1u. . ;,-g nonh~lmqn gene, l~ ;pt;oll of a cDNA minigene typically is driven by a linked gene promoter and enhqncer from the n~lr.qlly-Oc~;!lll;l~g gene.
As used herein, the term "large l.,.nc~ ç~ or "large homologous 25 ~eling construct" genP,r.qlly refers to polynucleotides that are larger than 50 kb, usually larger than 100 kb, frequently larger than 260 kb, occasionally as large as 500 kb, and somPtim-Ps as large as 1000 kb or larger.
As used herein, the term "I.,.ncr.;lltional unit" or "l~nc~ ional complex" refers to a polynucleotide s~U~Pnce that comprises a structural gene 30 (exons), a cis ~cfing linked plu",oler and other cis-actin~ se~ Pnces nP~ for efficient ~ )tion of the structural s~uenre~, dist. l regulatory e1P.m~P.ntc 15.
n~c~s~-.y for ayp~pl-ale tissue-specific and developm~nt-ql trqn~c~ n of the structural s~ ces, and ad~litinq1 cis se~ onres illlyGl~l for efflri~ nt I .,..-s- - ;yL;~ n and tr.qn~lqtinn (e.g., polyadenylaLion site, mRNA stability controlling s~-.lu~
Generally, the nnm~nrlqt~lre used herein and the lal~ld~l~ 5 y~ ,s in cell culture, mc'~clllqr genetics, and nucleic acid c~ and hyl~ ;m~ de3c~il~ below are those well known and commonly employed in the art. St~ h~-iq~es are used for l~o...h;n~ nucleic acid meth~l~, polrll~leoti~e sy.~h~s;~, cell culture, and ~ g',nf, il~COlyOlaliOn (e.g., liyore~;lion pl~tocols). fiPn-or,qlly e~yllldlic re~ctinn~ oligonucleotide ~y~lhf~;~, and 10 pllrifirqtinn steps are p~.rollllcd accoldi~g to the mqmlfi~c~ 's ~ ;rA~;on~ Theterhniql~es and pl~lul~,s are gcen.orq.11y ycl~,lllled accol~ding to coll~f-nl;nn~l m~.th~s in the art and various general l~f~ces which are provided I~Jugh.Jul this ~lo~...rMI The plucedul~s therein are belie./ed to be well l~own in the art and are provided for the coll~,eni~ of the reader. All the i~v.. ~l;OI~ co.~;n~1 15 therein is incollJolaled herein by l~,f~ ce.
~ himeri~ f ge~e~l mice . re derived accol~ling to Hogan, et al., Manipu~7n'ng the Mouse Embryo: A Laboratory Manual, Cold Spring Harbor Laboldt~ ~ (1988) and Teratoc~i,uJ,, as and El~r ~ JI~iC Stem Cells: A PracticalApproach, E.J. Ro~l~n, ed., IRL Press, W~ h;l~ol~, D.C., (1987) which are 20 il~cGllJu.dted herein by l~f~ ce.
Embryonic stem cells are mqnir~ acco .lil~ to p~ hP,d pl~lul~s (Teratocarcinomas and Emb7yonic Stem Cells: A Prachcal A~,oa~h, E.J. Rol~ll~n, ed., IRL Press, W~cl-i~ on, D.C. (1987); yilstra et al., Nature 342:435-438 (1989); and Sch~.~t~lg et al., Science 246:799-803 (1989), each of 25 which is incol~lated herein by ,~f~ ce).
OligomlA1P~tides can be sy~ d on an Applied Biosystems, Inc.
(Foster City, CA) oLgo---Jc~ P s~..lh ~ accor~ling to .sl~e~ifi-~ticns provided by the mqmlf;~rvtvrer.
It has often been obsel~ed that cDNA-based ~ Cg~ fS are poorly 30 c~,l~scd or inapJ~lopl~tely regulated. ('Jennmir DNA-based ~1. n,Cje~ f~S (i.e.,consL~u.~ed from cloned genomic DNA se.luçnr,es) which sub.t~ l1y retain the 16.
content and o~ on of the nq'llr.qlly-oc,;.-. . ;.~g gene locus are more likely to be co,~ctly eA~ ssed, but are limited in size by the cloning capacity of b~a~ )hageand plqQmi-l/cosmid vectors. The yeast artificial ch~ol"osoll,e (YAC) is a l~ll11y developed cloning vehicle with a capacity of at least a~ vx;...~lPly 2 mP.~ahqQes (Mb) (Burke et al. (1987) Science 236: 806). An upper size limit has not been 5 ele ...;,-~ to date. The ability to l~pn~lucibly and Pffi~i~P.ntly introduce YACs into gel-ic mice can Qi~ifi~qntly surpass current l-~ .SgellP. size limits. Meth~s used herein for YAC cloning, mqnirllqtinn and genetic mc(lifirq~inn are those well known an~l commonly employed in the art. See for çYqmrlP., United States Patent 4,889,806 and WO 94/00569. B. sic mpth~s of yeast genetics, inl~ln-ling 10 desc,;l~lions of SP1~tq~lP ...~ and selective media, are found in Meth~Q In Enzymology (1991) Vol. 194, ~ lemi~ Press, Inc. ~d~litionq-l mPthn~Q can be found in Current ~olocols in MolP~ulqr Biology, ed., Ausubel (1994), Greene Pub.Associàt~,s and Wiley - Tntp~rs(~ip~n~e~ J. Wiley, New York, NY, particularly volume 2 chapter 13 "Growth . nd M nipulation of Yeast"; all volumes of which are hereby 15 inco,~,aled by ,~,f~.~ce.
It is an object of the present invention to provide methods for constructing laIge DNA constructs as Yeast Artificial ChromosomPs by homologous recombination bGlweell YACs (lcÇ~ ,d to as parental or input YACs) during mPi~ sic in ye. st. In genP.r.ql, h~mologous recoml);~ n bGlween a pair of YACs 20 that share a region of homolo~y is o~l~ned from a three step p~cess: by first mating yeast cells of o~..ile mating types, where each type conl~;nc one YAC of the YAC pair to be ,~cc-..b;-led, to form a diploid that con~inc both input YACs, .
then sporulating the ~iploi(l~ i.e. in~lring the diploid to undergo meiotic division and haploid spore formation, usually by nill~g~ll starvation, during which 25 recombination bGt~ the YAC pair can occur, and subse~l~Pntly idcnlirying (and isolating) those haploid spores bearing a desired recombinant YAC. ~lthough the invention can be applied to recomhin~tion bGIweGn any YAC pair sharing relatively large regions of homology or l~lween a YAC pair di~Lyh~g mitotic stability when co-propa~t-P~ in a ~liplni-l, the mP.tho~c of the invention find particular use when 30 applied to YAC pairs with homology regions of less than about 50 kb, preferably less than about 40 kb, more preferably less than about 20 kb, even more pl~f.,..lbly WO 96tl4436 PCT/US9S/14966 17.
less than about 10 kb, and most p.Gr~ly less than S kb. The m~th~s of the invention find particular use when applied to YAC pairs that display mitotic inco...p~ihility when ~l~r...l~lc are made to co-propagate them in a diploid. The methods of the invention are particularly useful with YAC pairs having both limited homology regions as cles~-;l~l above and mitotic inca....~ ;lity. Accor~ gly, it is 5 an object of the invention to provide m~th~ls that increase the frequency of h~rloid yeast spores c~ ~; inh~g the desired ~o...l-in~n1 YAC after the sporulation step, and to provide m~.th~s that select pl~f~nenlially for desired harloi-l cells from the spore popll1qti~n to thus increase the frequency of desired cells within the population of cells to be analyzed for the p~e~ce of the desired recombinqnt YAC. Accor~ing 10to the i~ n, m~th~, which include steps for construction of ~lirloi(1~ prior to sporulation and steps for s"l!3~1uenl genetic so1~ti- n of YAC Co.~ ;Q;Qg spores, which steps can be used s~,p~.AtPly or, most ~.~,fe.~ly, in coml~L~n to best achieve the fo~goi.~g objects.
Accol.l~g to the invention mPthnfl~ for diploid constnuction are 15 provided such that during the mating step a diploid yeast cell, CoQI~ a pair of parent (or input) YACs that share a homology region suitable to enable homologous n ~,~n them duDng m~i~tsi~, will l..,;.n;.ii. both input YACs without ~g~ , It~, upon eQI~ ~ g the sporulation step. This is accomplich~A. by ~-~;--;---;~;-~g or ~ nling diploid growth, i.e. mitotic division (do~bling) of the 20 diploid cell, ;------P,~ ly after the mating, during the mating step. A rnlo.th~l of p~ ing a l~...~h;n~n~ YAC is provided that in~ l~des the steps of (a) mating a first haploid yeast cell co...~ g a first YAC to a second h~rloiA yeast cell Co..~ ;Qg a second YAC having a ht)mr~ gy region with the first YAC, to obtain a diploid yeast cell, wh~ ,in mitotic doubling of the diploid is limited or pl~ie.-~ed, 25 (b) sporulating the diploid and/or its mitotic p~geny, to obtain spores, and then (c) id~ ilying spores that c~..-I-.;.ce the reco...l-i~ l YAC. ~11. -..3~;vely the method - in~ Aes the steps of (a) mating a first haploid yeast cell compricing a first YAC to a second hqrloiA yeast cell caJI~ g a second YAC having a h~m-~10gy region with the first YAC to obtain a diploid yeast cell, (b) spoIulating the diploid and/or 30 its mitotic ~IUg~,.ly to obtain spores, and then (c) idenlilying spores that co..~I).;ce the recomhinqnt YAC by c~llh~fing spores for growth of spones co..~Ii. ;cin~ the Wo96/14436 PCr/uss5ll4966 18.
recombinant YAC, and optionally, sPl~cting against spores that comprise a parental YAC(s) and/or an ui~desilGd lGcoll,bi~nl YAC. When limited diploid growth during t-h-e mating step is desired (for example, as in the case when a particular YAC pair is mitoti~qlly ii~co...~-lil,lc), it is limited to about 8 or fewer diploid doubl;.~s, preferably to 6 or fewer d~ubli~gs, more preferably to about 5 or fewer, 5 even more pl~,f.,l~bly to 3 or fewer do~ e, or most ~l~f~.~bly p,~ ed.
Diploid doukling during the mating step can be delayed, slowed, limited or prevented by mr~l.;lo.;i~ the doubling during the mating step, by ~hPmirql means, by genetic means, by cU1tllring conditions that favor mating but disfavor mitotic growth, by 1;...i~ g the time of the mating event, by l;~ ;ng the time of diploid 10 growth, and by qltPring the te~ during growth. For GA~ll~le, mutant - h~rloi-1 yeast cells can be used whose entry into mitosis can be selectively controlled, for eYqmr1p by a change in lGlll~dtUl~,. Or for eY-q-mrlP, haploids having the same awuJLI~hy can be sep~ ely pre-grown in mPAinm providing the needed mltriti~ nql supplement, such that they would intPrnqlly a~ lqts the 15 mltritionql s"~lr ..l nl and then the cells are mixed lOgG~ for mating on auxotrophic media (lacking the ....I.;I;Ql~l supplement) such that the cells would have a suffiri~nt intP.rnql amount of the nlltritionql supplement to allow mating but not ~.~b~l~Pnt diploid growth (or at least growth). More ~ificqlly~ for e~ lr., YAC-bearing harloids, each leu2~, are sr,~ lely pre-grown in leucine-co..l;.;.-;.-g 20 mPAinm and allowed to ~-c~lm~ e lPu~inP, and then l-~sre.l~,d to leucine-lacking~P~ for the mating step. Drug s~silivily based control can also be used.
When mass ...~ g,e, (i.e. mixing of cultures of the parental h~rloid yeast cells, pl~ dl)ly l~tw~n 107 to 108 cells of each type are ~.Ço,.lled, diploid mitotic growth can, for e-Y~mrle, be readily limited by l;...;l;ng the mating step to a period 25 of time equal to or less than about 8 diploid-doubling times, preferably to 6 orfewer doubling times, more ~f~ably to about S or fewer, even more preferably to 3 or fewer doubling times, or most preferably prevented. By "mating step" is me. nt to the period of time co ..~.Pn~;ng with the miYing of cells of each mating type, inrlll~ling a lag period prior to schmoo fo.... ';~rl, and inr1ll-1ing diploid mitotic 30 growth that may occur when non-~yll Lo~ously mating cells are used. Diploid dollbling time under the con~ ;o~C of mating can be routinely dele ...inP~ by using WO 96/14436 PCT/US95/14g66 19.
sL~d~.l p~locols found in the art. Most, pl~felably mating is ~.rulllled in the absence of selP~ctir~n for ..~ s located on the YAC ~rms. This lack of S~P1Pction for t_e desir_d diploid runs coulller-inluiLive to est~h1i~hP~i yeast mating procedures.
As _as been Lscu~ ed by the present i~ ul~ applir~ti~n of genetic sPl~tion for ~irloids co.~ .;n~ both YACs, such as by sP~l~tir~n for .. ~ , on the YACs' 5 arms, s~ )ri~ingly resulted in the cul~wlh of ~lirloids in which ~ e~;.,.bl4P
gr....~nl~ of the YACs had oc ;ull~d. However, as taught herein, the cuullt~,r-ui~ive ~hspnce of a sfl~ n for ...~ on the YACs' arms prior to sporulation cil~iulll~ the ~lub~r~ caused by ~ - d~ - ~'~ ~.. ~,~... ..,l~.Alth~ugh 30~ is a le~ e typically used for yeast cl~lturi~, diploid growth can also be limited 10 dunng the mating step by p~.Çull l~g the initial period of mating at a higher , than the l~ ng period. The initial high te-. l~,.. t~ period inr,ln~les a period for the haploid parental cell to form a "scLlloo." The period can be ~t. ~..-;~-~d by ...o..;~u.;i-~ the cells, ~l~,f~.dbly miclûscopically, for the a~ re of "sc~ oos". Subs~lu~--nly, mating cells are ll~r.,.l-~d to a lower ~P",l~ , at 15 which cell fusion and conjugation conLi~lues but mitotic division is l~ ly l~led colll~ d to the higher telll~e.~lul~. An upper télll~dul~ for the initial period is limited by its loA,cily to the yeast cells. A lower l~ lulc is limited by its toxicity to cells or ~Jlevc;lltion of conj~ga~i~ n When desired a ~ler~
te...l,f,., l",~ ~ange for the initial mating period is about 25 to about 35 C, more 20 pl~f~ bly about 28 to 32 C, and most preferably about 30 C. A p~f~
te...l~,., l.,.~ for the l~ g period of mating is about lS to 25 C, more preferably about 18 to 22 C, and more pl~f~.~bly about 20 C. In one ~)er.ifi embodiment of the ill~G~tion mating is mr~n;lo.~d for several hours to (; v~
such that mating is complete, or nearly compl~t~, but diploid growth is limited to 25 ~ro to three generations as follows. A~pl~ ely S x 10' to 108 cells of a MATa strain is mixed with an appl~ y equal .. --..be~ of cells of a MATa strain, vortexed, and bulk plated as an applu,.i~ p~ly dime-si~d drop onto a 10 cm YPD
plate. After four hours at 30 C, the patch is evenly spread out over the entire surface of the YPD plate and kept at room to-..l~ d~ c (about 20 - 25~c) overnight 30 (about 18 - 24 hours). This particular l~lVoedul~ results in effici~nt mating with appl~ lely one to three ge--e .,l;ons of diploid growth. One can readily adjust WO g6/14436 PCT/US95/14966 20.
the time periods, te~ .e or other con-lhi-n~ con~i~tent with the mating and diploid growth rates for a particuhr combination of h~plni~
Recom~in~ti-)n (meiotic l~...h;~ inn) can then be in-luce~ by sporulating the ~iirl- id S~u~ sporulation lUUI- es can be used, with t_e proviso that further diploid growth is not e~ ge~ The res~llti~ sporuhted culture is S
then Pn.;-hPd for spores using known p~tocols. nn.;~l.,..Pnt will result in a ,subst~nti~lly pure spore popul inn. By s~ ;ally pure is meant about 90%
spores to about 10% ~iirloi~s and ~ ~ h~rloids plug~ssing to a more ~ f~
about 95 % to 5 %, to an even more p.~,f~ d about 99 % to about 1%, to ideally about 100% spore population. ~lthnllgh any ~tand~l sporulation pl~tocol can be 10 followed, a plGf~l~d pl~tocol in~ les removing about 10~ cells from the mating plate, ~ ing the cells in sterile water, l~"~l~.n~ g the washed cells in 10 mls of sporuhtion ...P~ ..., and i~ g the cells at 30 C for three to five days.
Several 1;,~1;ngs l~dnding sporulation effil~iPnry ~ d from these t;~ P'~
First, sporulation effi~ ien~y drops off ~lranlqtirqlly when the culture density e-Y~eeA~s 15 5 x 107/ml. SPccn~, ...~;.~1Pn~nfie of sporulation cultures in polyethylene tubes ~ cLiv~ly precludes sporulation. Best effi~ :~.nc:f s were obt~ined in poly~lyl~ne or glass tubes. A pl~_f~ d sporulation ...?A;~.. iS 1% pctq-~ci~lm acetate s~lpplP.mP.ntPA
with one-fourth of the normal qmollnt~ of amino acids (Rose et al. (1990) Methods in Yeast t~Jen-Pti~ s A Labol~.~ly Course Mqmlql, 1990 PAiti~n, Cold Spring Harbor 20 Labol,l~ly Press, which is hereby incol~ldted by l~f~,nce in its e.~ y.) qlth~ugh other standard yeast spon~lation, particularly for S. cerevisiae, wi11 suffice.
The spores can be s~ ed by the usual mPth~lc in the art for the desired recombinant YAC, or p~f~.~bly, a method of genetic selection acco~du~g to the invention, as ~ senled below, is followed in which case step (c) in~ les 2S
id~ ifyil~g spores that collll,l;se the desired l~CGlllbl~ YAC by cllltllrin~ spores under con-litil nc th~t select for growth of spores CO~ ;S~ng the recombin. nt YAC, and opti-n~qlly, that select against spores that compnC-e the ulldesiled re~ombinant YAC and/or one or both un~llll)~ed parental YACs in order to select for and enhqn~e i-lPntifi~vqti~ n of spores cc.~ ing the desired YAC. 30 M~P.th ~s are provided for the spore analysis step that fq-ci1itqfe ~letectinn and i-lPntifi~qtion of spores having the desired recombinant YAC by 6 ~CT/US95/14966 21.
~,rO~ g a genetic selPcti~ n for yeast spores con~i~;n;Q~ the desired recombinant YAC, while optionally cou,l~er-selP~ g (i.e. selp~ting against) for the undesired by-product recombinq-nt YAC and/or one or both parental YACs. Genetic s~Plection iscl~P-ci~n-P~l based upon the ~ e~led s~P~hl-P ...z.k. ~ ~ present on the desired,Gc~...hi.-~..l YAC in cc.l.p~.;.coll to those present on the b~l~luct l~".h;nlnl S
YAC and/or one or both parental YACs. Acco~ing to the invention, genetic s~1Pc~ n is p~lr~ lled by providing the input YACs with sP1P~t-qble "~a,k~ -~ such that YAC recoml~ n results in the desired recomhinqnt YAC pr~lirt-q-hly having a certain S~PlP~t-q~ P- marker, orpreferably ...~.k -~, that allow applirq~inn sP~
con-liti--ns for the marker or ...~.k. -~ to obtain pl~f.,.~ l growth of spore pl~g~l~y 10 co--1; ;,-;~ the desired YAC in the pl~,~nce of spores cn.~ ing a YAC having a - dirr~ s-plect-qhle marker or .. ~ , and optinnqlly, to select against the marker or ~a~k~ ~:i present on the undesired recomhin-q-nt YAC and/or one or both parental YACs. After diploid COnStluC~Il, ~l~,f~ably by a mP.thod of diploid constructionacconling to the present invention, and sponllq~i~n~ spores are cn. ;rh~ with respect 15 to ~l;l)loi~lc and un~ ~I h~l)lhirls by mPth~lc known in the art, and the e~ h~lspore population is s~ect~P~ to genetic selection in media that selects for the desired l~o...l-inq..l YAC and, opti~nqlly but preferably, cuu"t~l-selects for haploids bearing the undesilcd recombinant YACs and hqrk~i-ls bearing either parental YAC. The s~ n step can be desig~ 1 to select for at least one 20 sPle~-q~' marker present on the desired l~o--.l~ YAC, and more preferably for at least two s~ hle ~r~ that are present on the desired ~co---h;~n~ YAC but that are not present on the undesilcid lcco.--binq-~1 YAC. More pl~f~ably cùunl~r-sele~ion is also ell~plo~ed for cells with either the undesilcd recomhinqnt YAC and cells with one of the parental YACs. Most preferably, countc.-s~PlPcti~ n results in 25 cell death. Any s-P~Pctqhle genetic marker or combination of ~r~ can be used so long as a genetic s~ ;t;ol- can be ~e.Ci~-f~ for the marker that selects for growth of cells cc ~ inil~g the desired YAC, and so long as the genetic s-Plection for each ma~ker is not mlltuqlly exclusive when used in comb~alion (or with a count~r-s~P1P~ir~n). Au~ollvphic .. ~.1....... ~ a re pl~f.,.l~,d for ease of use, although drug 30 , metal (toxin) tolerance ...~ and the like can also be used.
Commonly used, and pl~,fe.lcd, .-.q.l~ in yeast are URA3 j~RPl, LLU2, LYS2, 22.
ADE2 and HlS3. Preferably, although not necf sC-. ;1y, marker scle~;lion occurs COllCo i~lly, rather than seqllPnti~11y. M~rkPr~ that find use in counl~-S~P1P~i-n are those known in the art and include URA3, and LYS2. I~fe~Yed is the URA3 marker in which 5-fluoro-orotic acid ("FOA") supp1-Pm-Pnt~ti~n to the media (Boeke etal. (1984)Mol. Gen. Genet. 197:345)resultsinthedeathofcellsbearinga 5 genolllic, pl~cmitl-borne, or YAC-borne wildtype URA3 gene. Also pl~cÇ~ d is LYS2 which also has a co!~n~ electi~ n scheme. The chrl~ 1 alpha-amino adipate is used to select against LYS2, as the LYS2 gene product pl~Jcesses it to a toxic product. Yeast h~r~ l strains preferably have non-l~ ,.Ling ml)t~ti-~nc, .e.g.
disluyl~d null .-.~J~ n~ in the l~ f,~S of interest in order to ~--;n;---i~ bac~g~und 10 false yosilives. Most preferably it is desir~l lP not to have s.-l)pl~ssil~le or l~ Lil le mutant alleles in the ba~ unds, since there will be false po~ilives coming from S.lyplf SSOl~ and, less likely, l~ . Nom,_~f,~ g alleles like de1Ptionc/insertions are pl~,f~ble to suy~ il,le alleles like amber or ochre ".,.I~;onc or r.~...f~.cl.;rl~ As a col .~e~ e of the mP.th~ of the i"~enlion, the 15 resl-1ting surviving cell popll1qfion _as an incl~d frequency of h~rloitl cells co.~ ;ng the desired YAC, and accoldinE,ly, the s~s~ eM~ sclo~nil-g and analysisfor cells l~l~ling the desir_d l~ll,l)ina.lt YAC is greatly .~imrlifiP~l and c~l)eA;lr,d If the input YACs ex_ibit mitotic stability in the parental _aploid 20 st~ins, the strains are typically eYpqn-lPd in sele~live liquid ...~A;--... (s~lec~ivc for the input YAC), and only one clonal population of each parent need be used for mating. In the case where an input YAC (or both input YACs) may eYhibit some mitotic in~qhility in the parental haploid state, mllltirlP. s~clo~s of each of the two haploid YAC strain can be grown non-selectively to a large patch with s~ll s~lupnt 25 mating by miYing a fr~~tion of each patch to all possible mates. Then, if desir_d, while the cultures are in spolulation ...P~ ...., the parental (haploid) s~ nP~ can be s.;l~ned by pulsed field gel electlv~lc~is to identify the subclrnP, of each mating type c~ ....ni...~l YAC loss. Purified spores from the cross bciweenthese two h~rk~id subc1~ neS are chosen for the genetic sPlPction step of the 30 invention in which one selects for the recomhinqnt YAC and optionally against the undesired by-product recombinant YAC and one or both parental unrecombin~nt WO 96/14436 PCT/US9~/14966 23.
YACs. Note that the % YAC loss of a subclone patch is a st-q-ti~ti~-ql p~ y of the patch and not a genetic pl.~lly of the subclone, i.e., a low % loss patch will display a di~llilJuLion of % loss in its subclones. The dislli~Julion of % loss is ~le~f nr1~ -~1 on the mitotic stability of the particular YAC. A veIy lm~t~b1e YAC will require S~;lcen~ng of many more ~.,bclnn~s than a stable YAC to identify a patch 5 with sllffi~i~ntly low % YAC loss.
~ lthnngh either the step of ...;i~;...i,~l;nn of diploid mitotic growth or the step of genetic s~,lection, pl~f~ldbly sPl~l;n n for the desired ~ecombinqnt YAC
and against the undesi cd YACs, can be pe.ru,llled without the other as a step in a known YAC recombil--~;o n protocol, in the most l~lcr~ ,d embodiment of the 10 invention both steps are pclr~ lled. Con~~ .-lly, lcco...bil-~;o n of unstable YACs comprising short regions of hnmoll-gy can be more readily ob1;~;nrA by the present invention.
The invention finds particular use in constructing large DNA
constmcts of l-~sg~nes and large homologous ~;cLing constructs .sl)Anning at least lS
one comp]ete ~ .;pl;nnql compl~Y, suitable for L-~r~. into ~ liqn cells, such as elllb~yonic stem (nES") cells, typically for construction of a ~ sg~,~icanimal.
Um~langed ;...~ nng]~Ulin genes cloned in YACs can be ~luced into ES cells and d~ ~loped to form a ~.,.n~g~ni~ animal in which 20 pr~ductive VDJ l~ -e~ ..- -.1 occurs, and tiA~l~S~ioll of i~ -nJglobulin chains also occurs. Large llansg~lles can be cloned in YACs and, after isoLalion from the host yeast cells, effil iPntly ~ r~.l~ into ~--~-.--..~li~n cells (e.g., ES cells) wilLoul prior sepalalion of the desired l~ e s~P~u~pnre~ from yeast-derived YAC
se~uPnees, and that the pl~ce of such yeast-derived YAC s~luenres can be non- 25ult~ r~ in~ (i.e., co..~ le with effi~i~Pnt l-,---~g,~ -leg...l;l~n and ~ r~ ion of a I ~ s~l-e 1~ ional unit). The present m-oth~s may also be carried out with - som~tir cells, such as eFithPli~l cells (e.g. j -ke~tinrJcytes), entlothrli~l cells, h.~ opoietic cells, and ~ly~yles, for eY~mrl~-. A large ~ g~c.~p can be nnnhnmologously ;~-te~ e~l into a cL~.. oso.. ~l lo~tir~n of the host genome. 30 ;vely, a homologous ~lhlg construct (which may colll~.lise a l.i~nsgene) that con~ins at least one altered copy of a poItion of a gerrnline gene or a 24.
xenogenic cognate gene (in~ tling heterologous genes) can be introduced into thegenome of embryonic stem cells. In a portion of the cells, the introduced DNA iseither n-~nhqm~logously ;III~ Al~l into a chromosomal location or homologously recomhinPs with the en-logPnnlls (i.e., nq1nMlly occllrring) copy of the mouse gene, ~;ng it with the altered construct. Cells co.l1Ail-;ng the newly ç~g;l~.~l 5 genetic seq~l~Pn~(s) are injected into a host mouse blastocyst, which is l~.;...p1-nlP,~
into a recipient female. Some of these ~ bl~os develop into ch;...---;s mice that possess a poplllqtinn of germ cells partially derived from the mutant cell line.,ro~, by bl~ing the rhi~.. ;c mice it is possible to obtain a new line of mice COII1A;n;~ the introduced genetic lesion (reviewed by Capecchi et al. (1989) Science 10 244:1288, ~col~ldted herein by l~fe~llce).
For hnmol-~gous lalE,~ g constructs, l~g~ting effi~iPnsy generally in~ ,ases with the length of the lal~ ing llal~sgel~ portion (i.e., h-)m logy region) that is s~ Anl:~lly co '~ nr.~A.~ to a l~f~ ,nce s~u~nce present in the target DNA (i.e., clui,so~e~ target s~u~~ e). In general, ~E,eLing eff1riPnsy is o~li",i~ed 15 with the use of icogP.ni~ DNA holllGl~ regions, although it is l~og~d that the p~se.ue of lecolll~inases in certain ES cell clones may reduce the degree of se lu~nce ;dentity l~ui~ for effi~iPnt recoml~;n~ n T~AnCg~Cn~S which encode a gene product that is ~nog~C~-ic (e.g., heterologous) to a n~ k~ n host species are useful. Such I~AI-Sg~C- neS typically 20 c~...l..;~e a sLIu~:lul~l gene s~lu~---ce ~ )lessiûn cqcc-P,tte, wh~..,;n a linked promoter and, ~l~f~bly, an ~n~ drive C~ Si~i~n of structural S~1U~ ~S encoding a IrPno~ ~ic (e.g., ~t~l~uus protein). The polym~clPoti~le sequ~Pnce encoding the xenogenic (e.g., heterologous) protein c n be operably linked to cis-~~ting I~AI~C~ ;1d;~n~1 regulatoryregions(e.g.,promoter,enhqn~P,r)sothataheterologous 25 protein is ~A~l~sed in a llla"ner similar to the t;~lcs~ion of the cognate endog~nuus gene in the n~~l-rqlly-ocul...;ng nnnhnmqn animal.
The presellt invention thus finds particular use in constructing YACs, genPrqlly by l~---h;~ ;on of YACs obl~ led fïom genomic lihrariPs but also by recombination with a gen~ti~qlly e~g;nr,-ed YAC ~lecign~ to provide desired 30 I.. ~.. ~liqn cell selectable .. ~ , t;A~l~s;,ion regulatory regions, protein fusions, larger regions for homologous ~~e~ing, and the like.

CA 02204359 1997-0~-02 25.
The meth~l~ of the present inventlon find particular use in constnucting minigPnes th~t comprise clusters of genes that are not ncrmgl1y found clustered in a genome or that are located in dirrGl~ lL gel~o~ ~. In many ;.~ s single ccmpo~G~ls of a multiccnl~onGlll complex, a metabolic paLh~ay~ or a regulatory- p~tL-.ay~ e.g. i.. unon.~h.1~ti-n, are encoded on se~ P genes, but all 5 of the c~ on~nl~, and thus all of the genes, are needed or desired to enh~nce orobtain fun~ti~n One such group would be human cytokines which find use to support the development of ~ "~ 1ed human he,nâtol)oietic cells in a mouse. Of the dozens of ~L~ es c~ d only a few are n~hl~lly clustered. Prior to the meth~s of the present ihl~e~l~ion, typically each cytokine gene of interest for 10 ob~;ning desired rwl~,1ion would be used to create a s~-...le l-, n~, ~;c line, which lines would then be ,llt~ .l.wed to obtain a desired line having all of the cytokine n~e ~~ of interest. In ~iti-~n to the time and labor ~u~ed to concollliL~tly genf-~te, select, and ",~i,l1 ~;-, â mouse line(s) for each gene, aw~ ely one year of ih~ ing would be 1~4u~h~d to obtain the f~ desired line (in the case 15 where about S genes were used). ~1th-n1gh it has been reported that coil-je~l;n~- of s~.Ate genes (up to at least S genes) may result in ;..t~ l;on at the same site (and thus be gc.l.~ 1ly linked to each other), this approach has problems. First, thec~ ted l~n~g~cn~ aIrays would be u~ liclable and ul~collllvllable in ~U~;IUl~. Con,s~u~ 1y, the stoiC1~;o----~ of the co,ll~onf~l genes of the Llansgelle 20 array could not be controlled, and in -1~1itinn, ~ bSf~ Jp-d structural analysis of the ge--, array would be very .1iffii~ 1t if not ill~ sil/lc. Second, it is desirable to use ~no~.;c f~gmP.ntc of snfficient size such that each gene is ~ 1;rq11y regl~1qted That is, by adjoining small 1~'n~g~nf~S (e.g. less than 30kb) it is highly likely that they would affect each others' regn1qti~n. By adjoin,llg large genomic 2 f~rn~.ntc, e.g. about l00kb, e. ch gene would be better burr~,d from each other with respect to ih~l~Ç~i,~g with gene re~11qfi~nn. Genomic Pl clones (Callj~ng inserts of about 75-lOOkb) are an ideal source of such large gt~nnmic frqgmPnt~.The Pl cloning system is based on a modified Pl phage replicon as vector. The salient feature is tllat the insert size range is about 75-lOOkb. Also a SC~.lil~g 30 service for a human genomic Pl library is available (-G.o,nnm~. Systems, St. Louis, MO). If Pl DNAs are microinjected, restri~fions on the amount of DNA which is WO 96/14436 PCTtUS95/14966 26.
micl~ ;nje~i.hle would limit the nllmber of copies of each gene in the micl~;njP~t DNA ~lulc. On the other hand, the cloning capacity of YACs p~. ,ls the construction of a single l.,..-.cgr;-P conl~;nin~ all the cyl~ c genes of interest, at least 5 in this eYqmplP. Using the mPthocl of the present invention, one would first obtain Pl clones (a~ lely 100kb each) each ch..~;n;ng one of the genes of 5 interest, convert these Pl clones to individual YACs, construct "linl~ng" YACs conl~h~;ng inserts derived from a small frqgmPnt of DNA from each of two individual genes or regions to be q1joinP~1, and then applying the h~m~-~1ogous recoml);~-~';on mPth~c of the present ~ ion to se~uf~n1;~11y recombine YACs in the desired order to obtain a final recomhinqnt YAC cnmrri.cing the desired 10t~ gf ~e array. T in~ing YACs aIe employed by first leco~h;l~ g by the mPth~
of the present i~,nlion a linking YAC with a first YAC co.ll;~;n;n~ one of two genes of interest to be adjoined. A linking YAC C~ ;n~ as its DNA insert a firstsmall region that is homologous with a portion of the first gene of interest and a second small region that is hnmQ1ogous with a portion of the second gene of 15interest. As a con~e~upn~ of the homologous rec~mbin-q-ti~n l~l~ the linking YAC . nd the first YAC driven by the first small region of homology, the desiredfirst recQmhinqnt YAC coll~ins the first gene of interest plus the second small region of homology (to the second gene of interest). Con~equently, a second homologuus ~CQ. .h;~ ;- n, l~t~n the first recomhinqnt YAC and a second YAC 20 that cont .;nc the second gene of interest for adjoi ~il~g, will be driven by the second small region of h-lmology, and by using the mPth~s of the invention will thus result in a obt;-;n;ng a desired second ,e~ .;n~ YAC in which the first and second genes are adjoined.
The DNA inserts for lin~ng YACs can be con~nie,llly constructed 25 in plasmid/ph~g~--.-;rVcosmid systems well known in the art using ~t~danl mo~~c~ r biology ~P~~h~ ues. Since these co"~nient systems typically allow m~nirllAtinn only of ~ldlively small fraEmPntc of DNA, the linl~ng YACs will in turn contain small inserts; the inserts of the linking YACs will thus only share small regions of homology for l~...h;.~l;.m with larger YACs that contain the genes of interest for 30 adjoining. Co.-.c~ ue~ly, the me.tho.1~ of the invention find particular use when linl~ng YACs are employed, since the mPtho~s enable obl~ illg a desired 27.
recombinqnt YAC res 11ting from recombination of YACs sharing a l~laliv~ly smallregion of homo10gy as taught herein. Of course, the m~th~ find particular use inrecom1~ ;o~- sch~mes that typically result in a relatively low recombination rl~u~,l ;y, such as when a mn1tir1~, preferably three-way, recomhh--l;ol- event is ~tl~ ted or when short homology regions are relied upon for homologous 5 recombination (as when using linking YACs).
~th-uls of the ~cilltion are provided for the construction of linking YACs by long-range polymerase chain reaction ("LR-PCR"). LR-PCR allows for ~mp1ifir~tion of a h- mo1c~y region obt~ined from a first YAC such that the ~mrlifi~ DNA can be ligated to a second LR-PCR ~mrlifi~d homology region 10 obt~in~ from a second YAC, followed by li~ti- n of the adjoined DNA to the arms of a YAC vector to create the linking YAC.
The vectors co~ h~ing the h~mo10gy regions for adjoining in order to construct a linking YAC can be obt~i~ed by ~~igestion of the first and second YAC
DNA at a ,~s~ ;o.- site which occurs only once in the vector ann, e.g. XhoI and lS
NdeI for the C~,lL,u~ ;c arm and XhoI and EcoRV for the acen~llleric arm. The tligestion products are ligated under con.liti~n~ prmting circ~r7~ton of the DNA f~gml~.nt~. The 1iga~i- n products are introduced into E. coli by electl~ tion orany otherconvenient l,~sro..~ ion method. The c~nl~ lic vector can be isolated by se1ecti~ m for qmpici11in ~ C~n~. If the YAC is cloned 20 in a YAC vector c~ tS-h-;ng a ~yCill ,~ marker on the ac~ ulll~lic arm (such as pYACneo), the ~.~ C,liC vector can be jc~1qted by sP1ecti~ n for y~,hl l~ t~n.~. Other s~1~P~;~-" sc1-P-.-Ps can be used as are known in the art and as are a~plùpriate for the ...~ on a particular YAC vector. ~l~r~ iv~ly~
the first and second YAC DNA 1ig,qti-~n products can be used di~11y for LR-PCR 25 i~n of the hnmo1ngy regions.
The p. ;...-, ~ for LR-PCR isolation of YAC homology regions are deci~l~l to hybridi_e to s~u~--ces near the ends of the YAC vector to ...h~i...i~e the ~..u~ 1 of YAC vector s~lu nr~s in the amr1ifit~qti-)n product. Each primer conCi~ of a 3' region of p,~f~l~bly apl),o~ P,1y 20 bp c~ inin~ h-~mc)1Ogy to 30 the YAC vector alm and a 5' ~egion ~l~f~,ldl~ly of ~ylu~ ly 20 bp con~;~h-i-at least one ,~ - site. The first and fourth l,~ h~ (for example, S3CX and 28.
SlAR l~e~lively in Figure 9) have at least one restriction site in common, and the second and third p. ;...~ ~ (for example, S2CRl and S2'AXl ~sl ecLively in Figure 9) have in common at least one ~s~ ;on site which differs from that found in thefirst and fourth prim-o.rs The restrir-ti-ln site of the first and fourth plilllC~ is chosen such that it does not occur within either homology region and is found as a 5 cloning site of the YAC vector that comrri~es the linking YAC. The restrirti~n site of the second and third primers are chosen such that it does not occur within either homology region or within the YAC vector arm se~uenr~s. ~ ;v-ely~ the rçstrirti-n site se~lences can exist in the ~.;...c-s as pre-cut regions with a~"o~ e o~clh~g for su~se~uent ligation. The s~lit~hlP size of the primer region bi~ding to 10 a vector for use in LR-PCR is well-known known in the art, but preferably is at least about 14 bp, more preferably about 20 bp, to about 35 base pair. The primer regions co.l~;..;ng the primer restrir-ti--n sites used for subse~nent fr~gmrnt ligation and liga~ion to the YAC arms, need not be, are likely not homologous to the vector se~ re but are rather syllll~lic regions des;gnfd to harbor the ap~,vp,iale lS
,~s~;rl;~n sites.
The first and second plihll~.s are used to amplify the homology region of the first vector and the third and fourth plhlle~ are used to ~l~lily the homology region of the second vector. The desired ~mplifir?tion products are ulirled by gel el~vph~,~,i,is and/or ethanol p,~ipildlion and tiigestçd at the 20 collllllon ~ ;o n site of the second and third plilll~.~ with the a~p,opciate l~s1 - ;~ n ~yl-lc. The digestion products are purified by ~ )aldtiVe gel ele~ hol~ s;s, mixed at a 1:1 molar ratio, and ligated. The ligatir n products are then digested at the restrirtinn site of the first and fourth primer, and the lligçstinn products icolqtP~i and yulirled by pl~ ive gel elect,~,yho,~s;s. The digestion 25 products are p,~,dG....n~ y cs...y.i~e~l of three ~imPrir DNA fia~n~nt~: the desired heterodimer wL~.~.n the dc ~. llsll~ll end of the first homology region is ligated to the uysll~l end of the second homology region, the hnmotlimPr of the first hnmology region and the hnm~limP.r of the second hnmo~ngy region. The DNA
products are then ligated to YAC vector arms at the cloning site co...~ ;l.le with the 30 restrirfinn site of the first and fourth p,;...,-~, and the ligrq-tion mix is introduced into a yeast host strain by yeast l~ ;on The desired linking YAC cQ~ ining the WO 96/14436 PCTtUS9Stl4966 29.
~t~.~li,ll~r in the same vector arm oriP~ ;ol~ as the parent YACsisidP-ntified among the llal,sro~ t clones by relstri~ ff- n digest so~ .. analysis. If the two homo1ngy regions differ in size, the heterodimer can be i.colqtP~d away from thehnmo(1im~rc by p~aldtiVe gel elecl-~",hol~sis prior to li~Ation to YAC vector arms, c~ h;ng for the desired linking YAC. Figure 9 illnstr.qtec the use of LR- 5 PCR for linking YAC const~uction and provides one specific embo limpnt ~lth-~ugh Notl l~s~ ;t)n enzyme sites are e~ l;lied herein, any other rare restri~tinn enzyme site can be used, more pl~f~ bly one that does not appear in the human genome, with the ~ viso that it does not appear in the YAC vector arms or in either homology region. ~lthnl~gh i-SceI is exemplified herein, any other rare 10 restrirti( n enzyme site can be used, more pl~,f~lbly one that does not appear in the human gPn~ .c~ with the proviso that it not be present in the YAC vector arms oreither region of hr~mology.
Large polym~cleoti~es are usually cloned in YAC vectors. For eYAmple, human genomic DNAlil,l~ies in YAC cloning vectors can be s~;l~ned 15 (e.g., by PCR or labeled polynucleotide probe hyl~ n) to isolate YAC clones sl.Anni~g complete genes of interest (e.g., a human APP gene, a human ;..------noglobulin heavy chain locus or light chain locus), or .~ .;r~ nl portions of such genes which col~,l)lise a co..-l lPte ~ t;onql unit. M~-th~c for mqki~
YAC libraries, icolqting desired YAC clones, and ~uliçying YAC DNA are 20 desr-. ;l~ in the art (IJ.S. Patent 4,889,806; Burke et al. (1987) Science 236 806;
Murry et al. (1986) Cen 45:529, in~ol~ldted herein by l~f~.~nce).
Genes and DNA regions of interest, as well as YAC vectors, for use in the mPth~s and co""~osilions of the present invention include those reported in:
Lonberg et al. (1994) Nature 368:856-859); Chen et al. (1993) Tnternqt Tmmlln~l. 25 5:647-656: Taylor et al. (1993) Tntrrn~t T...~ .nt~l. 6:579-591; Choi et al. (1993) Nature t'~en~otir~c 4:117-123; Chen et al. (1993) Embo J. 3:821-830; P~on and - Choi (1993) Proc. Natl. Sci. 90:10578-10582; and Taylor et al. (1992) Nucl. Acids Res. 20:6287-6295, as well as in application United States Se~ ulllbe.
08/148,177 filed November 5, 1993, and application United States Serial number 30 07/900,972, filed June 18, 1992, all of which are hereby il col~o~ d by lt;îe.~.lce.

30.
These same lef~ ces provide and discuss .~1h~ls useful for the introduction intohost cells of the large DNAs produced by the m~.thoti~ of the present invention.Once a desired YAC clone is isolated, and preferably depl~le;l.;,~d yeast-derived YAC s~.,el-~es may optionally be co,llplctely or partially removed by digestion with one or more re.stri~tion enzymes which cut outside the desired cloned 5 large 1.~ & se~uence; yeast-derived ~~uen~es are s~ ed from the cloned insert s&~u~ s by, for eY~mp1~, pulsed gel elec~hol~is. E~ti~ably7 a complete ul~ g~ YAC clone is used as a large l.,.,-~e~-~ or large hnmo10gous la~Gling construct in the m~th~ls of the invention.
~f~ YAC clones are typically those which cc plct-1y or 10 partially span large structural gene se~ .nces for e~unplc, human APP gene, human ;..-..-.~ globulin heavy chain locus, human immnnog1nbulin light chain locus, human al-~~ ~sin gene, human Duçh~.nne mnsc~ r ~ly~ ph~ gene, human ~.~ ;u~loll's chorea-~scoci~ loci, and other large structural genes, preferably human genes. 15 A lll0l1 P11 pYAC3 vector (Burke et al. (1987) op.cit., incol~laled herein by l~f~"ce), pYACneo (Traver et al. (1989) Proc. Na~l. Acad. Sci.
(U.S.A. J 86:5898, incol~lal~d herein by ,cr~ ce), and pCGS966 (Smith et al.
(1990) P70c. Natl. Acad. Sci. (U.S.A.) 87:8242, hlcoll)ulàled herein by ,~,f.,~ce) are YAC useful cloning vectors. 20 YAC clones co--~ c a large heterologous ~ n~g~nf find use, for eY~mr1P~ in lllelllods to llani,rer the large l,ansg~ne into a plwi~~ t stem cell line which can be used to pr.l~ n.~f.l~i~'. nol~h~ n ~nim~l~ following injection intoa host blastocyst. For example, WO 94/00569 reports s lccessful tr~n~fer into mouse ES cells of the human APP l.ansge,le carried on a YAC clone. Coll~tly 25 I;~-grled l~S cells are then Ll~li,r~ led into sui~l-1e blastocyst ho$s for gt~.n.-..,.l;on of çhimP.ric l.,.n.~F.niC ~nim~ acco,~ g to meth~s known in the art (C~peccl-i, M
(1989) TIG 5:70; Csl~hi, M. (1989) Science 244:1288, incorporated herein by ~cç~nce). Several studies have already used PCR to succe~sfully id~ Itify the desired l,ansr~;led cell lines (7immP.r and Gmss (1989) Nature 338: 150; Mouellic 30 et al. (1990) Proc. Natl. Acad. Sci. (U.S.A.) 87: 4712; Shesely et al. (1991) Proc.
Natl. Acad. Sci. USA 88: 4294, which are ,ncol~o,~ted herein by lcfG,~nce). This WO96/14436 PCr/US95/14966 31.
approach is very ~rrf; ~ive when the number of cells receiving exogenous ~;cL;nggf~f(S) is high (i.e., with ele~ )olalion or Lpofe~lion) and the treated cell populations are allowed to expand (C-q~ hi, M. (1989) op.cit., inco,~ldlt;d herein by ~cr~.~nce). The blastocysts co~ ;nin~ the ;~ çd ES cells are llowed to develop in the uteri of pseudopl~,g~lt no ~ n females and are born as çh;..~f~ Smice. The resll1t-q-nt I ~A~e~ ice are cl~ - ;c for cells having the large l.,.,-~erl-~(s)/h~m~ g"us L~ ;.-g constructs and are ba~hlossed and screened forthe p,~ ce of the I~A~genf (S) and/or YAC S~1UenCP~S by PCR or Suu~ blot analysis on tail biopsy DNA of Orr.~,-- ;ng so as to idenLiry lldnSgeniC _ ice h~ y~;uus for the ~., ~-~e.~ (s)/homologous l~ting co~sl,uc~s. By ~.rulllling 10the appl~l~le crosses, one can ~l~nluce a ~ ~ nic nonhllmqn animal holllo;Gyguus for mnltir1~. large t.,.i-~e~-f ~/homologous recombin-qti- n constructs, and ~ptionally also for a ~ E~ ~e e -~;nl a dirr~ ~ heterologous protein. Such I....-~erM;c qnimqls are ~qtiQf~to~ l~ ;",/..~1:.1 models for various ~ es linked to the Ll~uli~r~ d !.,.. -.~g~ne(s). 15 The YAC l~o,l,bm~lt products of the invention find ~llitir)nql use in the pludu~;l;oll in a host cell of l~cQ---k;n~ pl~lcins and recnmhinqnt multi-protein cr....~ ." and in the conco.~ nl pr~xluetinn of pl~tf;inS involved in a biological thw~y or other binl~rql pl~SS. The recombinant YACs of the i"~, nlion are useful for ~A~ SSi~n of heterologous l~ h;n~--l protein or pl~ s in a yeast host 20 cell (or other host cell ~ .Ç~,l",ed with a YAC or large DNA produced by the mP~tho~s of the i"~ tio~) for uses ;~ ;n~ ;~ub~e~luenl preparation and isolation of the r~comhinqnt p~Ot~.llS, particularly isolation of co p~YPs that previously would be pl~Gd by ~ P. G~ssion and ;QI ~1qt;-)n of the co,l,~Gnls from dirr~,.G~II
cells followed by in vitro q-QQpm~iy of the complex. Biological pdlhwdy~ involving 25 mnltirlP. pl~ ns or gene products can be l~const,ucted in a host cell by use of the large multi-gene DNAs that can be readily constructed by the mPth~ls of the hl~Glllion. Such host cells, such as YAC c~ yeasts, are useful as fq-ctcriPs to produce a ~,u~h-~ or products of the biological ~alllway of interest. In ~ ition~
the mpth~ls of the u~ tion provide large multi-gene DNAs that find use to 30 lccu,lsl,u~ a b -Ic~gi~ql pathway in the ll ~rolll,ed host cell in order to facilitvte 32.
the study of complex biological l~alll~ayi and their coll,~nenls by enqhling the use of powerful genetic and moleclllqr biology approaches, such as ml~tqtionql analyses.

EXAMpT ~
_xample 1. Construction of Input YACS. In this _xample the input YACs used in 5 the recomhin~';ol- r-~l~ ;...Pnt~ of P~.nl)1es 2 to 11 are provided.
a. Constmction of YAC Jl.3Lys. Figure 1 provides a partial n map of input YAC Jl.3Lys. YAC J3.1LYS (ap~ ely 108 kb) C(i~l~inS as its insert the SpeI-SpeI region of the w~ gP~d human ;... ,u~g1Obulin heavy chain locus ;.. --.i~ bulin (H) chain gene (WO 94/00569: 10 hereby i~col~lated by ,cr~c~ce) in pYAC Neo. This frq.~Pnt conlains at least one of each C~ 1~UilCd for correct l~langclllclll and cA~l~s~ion of a human IgM heavy chain m~lP~ulP. The fragmP.nt contains VH6~ the f~ln~tionql di~e~ y (D) se~ k~ , all SiA joining (J) segmPnt~ and the C~ corlsl~,l region se.E;...~
(~oflrpr et al. (1989) Proc. natl. Acad. Sci. (U.S.A.) 86: 5587; ~~ et al. 15 (1988) EA~BO J. 7: 727; Shin et al. (1991) EMBO J. 10: 3641). To pl~e)alC an input YAC having on its arms S~PlPCt~ ~ that can be used when in~ )n of a genetic selp~cti~n step is desired, the URA3 "left" arm (in this case the ac~ ulll~ic arm) of YAC Jl.3 (YAC J1.3 was i~ol~tPcl from a YAC libIary produced in the YAC vector pYACneol5; WO 94/00569) was moAifiP,d to become 20 LYS2+ (and con~c~luel~1ly ura3-) by di~lu~tion (Rul~ lei~, in Methods in Enzymology (1991) 194:281) of the left arm URA3 marker by ll~srullllil~g the yeast st~ain AB1380 (Mat a, ~+, ura3, trpl, ade2-1, canl-100, Lys2-1, his5; see Burke et al.
(1987) S~iP-n~e 236:806-812) that cn...l..;~es YAC J1.3 with the LYS2/neo fr~gmPnt from pRV1, a linear f~gmPnt co.. ~ g the LYS2 and neor genes fl~nkP~l by 25 (di~ ting) URA3 s~-~--ces (Sli~aSta~d and SchlPs~inger Gene (1991)103:53-59).
Ins~Lion of the dislu~t i~g DNA, in~ased the total length of the app,o,.;...~ ly 100 kb Jl.3 YAC to about 108 kb. Eight LYS+ ll~lsro~ t~ were screened for both the ura~ l/h~ty~ and for mitotic linkage of the LYS+ and TRP+ ph~o~y~cs; 3 isolates sL~.ing this desired genetic behavior were i~lentifi~l All 3 showed mitotic 30 stability of the YAC in the haploid (i.e. YAC l~t~ ;oll) to be about 70-80%.
Mitotic stability was e~ ed by a mitotic loss assay. A mitotic loss assay (see WO g6tl4436 PCT/US95/14966 33.
Current ~ulocols in Molecular Biology, Chapter 13, Ibid.) can be ~.ro~ ed by cu1~ring a yeast strain comprising a YAC in culture media that, preferably, selects for growth of the cells co..~ .ing the YAC, transfernng a portion of the culture to a non-s~ live, rich meAillm, typically YPD, con~;U;ng solid support (e.g., agar plate) in a way that allows clonal colony growth (typically "streaking" for single 5 colonies) of plGrc...bly at least about 20 single, well s~.~'~ colonies, more preferably at least about 50 col~-niPc, conv~ ~ly after about two days of ub~ion at about 30~C. Subs~lu -~11y~ each colony is analyzed for the ples~e or ~bs~nce of YAC. The analysis is conv~iGlllly pGlru, "ed by replica-plating colonies from the non s~l~;liv-e plate to a solid support co,-~in;ng me~ium that 10 allows selective growth of cells comrricing the YAC (i.e., a "dl~ul" ~f-A;.....
plate). After an ;m ub~ n period to allow for cell growth, typically 1 to 2 days at 30~C, the m~mhPs of colonies E;lU~. iilg on the selective ("dlo~ul") plate is d~ d and divided by the total number of colonies on the master non 3e~ ,1ive plate that were L~ r~ d via replica-plating to obtain a value reflPcting the " % 15 stability" of the YAC in the strain under the initial growth condiLiolls, i.e., an appluA;~ e pe-~~ ge of the cells in the culture that retained the YAC. Mitotic loss assays used to e~ e p1~cmi-1 mitotic stability are suit~le for use herein.
b. Construction of a parental h~rloid strain beanng YAC Jl.3-LYS.
Since one of the emboi;.. - .~1~ of the genetic S~ ;on step f~t~ p1;r~ed in the 20 ~-~-..l)lr-s herein (see below) was dec;gnf~ to require a h~rl-i-i stlain having a his3-genuly~*, whereas strain AB1380 is HIS3+ and his5- the Jl.3LYS YAC was ll~u,sr~ d into the desired genetic bae~lound by mating and sporulation using standard pl~tocolc. One of the ~sro,l. a,ll isolates (Jl.3L.6) bearing Jl.3LYS
YAC was mated to yeast hqrloid strain YPH857 (MatcY, ~6-, ura3-52, trpl-A63, 25 Lys2-801am, ade2-101, his3-~\200, leu2-~1, cyhR). After sporulation of the tliploid, tetrad tlic~ction yielded 2 spore clones out of 98 total spores that had - retained Jl.3LYS, were of the correct mating type, and had the desired his3- and ~SS+ alleles. These two clones were desig~qted Jl.3L6.13D and Jl.3L6.18C
(herein l~f~ ll~ to as "13D" and "18C", l~ ively). 30 c. Constluction of YAC YNN~y-HIS. Figure 1 provides a partial ~,s~ n map of YAC YNN~y-HIS. YAC YNN~y-HIS (a~lu~ lely 44 kb) 34.
co.~ as its insert a 10.5 kb (Nde-Spe fr.q~rnPnt) region of overlap with the 3' l~.. i.~s of the insert of Jl.3LYS (i.e. the C~ const~nl region seg.. ~ .l) adjoined to an a,?pro~ tely 18kb region co.~1~;n;n~ the C yl region and the rat heavy chain 3' enhqneP,r derived from p ye2 of Taylor et al. (1992) NAR 20:6287, which is i.~col~olaled by ler~ ce (see also Lonberg et al. (1994) Nature 368:856-859, 5 which is hereby incol~.aled by l~f~nce). YAC YNN~y-}IIS was constructed by i~olqting the 28.5 kb frq.~nP.nt co.~ ;ng the 10.5 kb Nde-Spe frqgTnPnt and the 18 kb C~yl/rat enl-5~ ,r frûm plqc~ y~3, then cloning the 28.5 kb frqgmPnt into thenot site of pYNN followed by tl~Çulllldlion into yeast strain YPH857. pYNN is pYACneo as de~;l ed by Traver et al. Prûc. Natl. Acad. Sci. (1989) 86:5898 with 10 the EcoRI cloning site l~ l cPA by a NotI cloning site. Y~y3 was constructed as follows: the 10.5kb NdeI-SpeI r.. ~.. ,1 from pJlNA (des~;l;l,ed in Choi et al, Nature rJenPti~s vol 4 pll7 (1993)) was end-filled with Klenow DNA polylll~lase and SalI l~nkers (New ~nglqn-l Biolabs Inc., Bev~,ly, MA) were added. After SalIgP~sti-~n, the frqgmPnt was ligated into the XhoI site of p~e2. Clones co~ nin~ 15 SalI r".gm~nt in the s. me l ,..-.~r. ;~t;o ~q1 o. ;~ l ;OI- as the C yl gene of p ye2 were i~L'MI;l';~, and the joined r".g~ ~nl ic~l~ted from p~e2 as an about 30kb NotI
fr~gmPnt This fr~gmP.nt was ligated to pYNSN vector arms and transformed into yeast. pYNSN was derived from pYNN by NotI/SpeI double digestion followed by NotI 1inlrP.rir~, res~l1ting in a pYNN d.,.;~,ali~e lacking the 613bp region ~w~n the 20 SpeI site and the NotI site. Eight h~srv...~ were picked and s~;l~ned by a .SouthPrn blot of XhoI digests. One I~Jrul~ t (des;g,-~lP,d y2-4) had the correct insert o.;. -~ l;nn by XhoI digest. This ll~sru,l~ was s~lbjeuled to further analysis by So~lthern blots (using pBR or the yl region as probes) of both mrligsct~P~ DNA on a C~ gel, and of l~s1.;l~;nn-cut DNA (XhoI, ~qm~, 25 EcoRI, SphI) on a coll~/en1;nn~1 gel. In all cases, the sizes of fr.qgmPntc were as e,~ To ~l~pa.~; an input YAC having on its arms sPl~tqhle 1~ kr. ~ that can be used when ;n.~1~.c;nn of a genetic s~ ;ol- step is desired, and in particular one that is co--.~q~ in comhinqtinn with YAC Jl.3LYS, the TRPl "right" arm (in this case the c~ ulll~;c ann) marker of YAC y24 was mQtlif1Pll to become 30 HIS3+ (and cnn~ nly trpl~) by disruption of the right arm TRPl marker by rolllling the YAC y2-4 host strain with a linear frqgmP.nt co~ ;u~ a HIS3 35.
gene fl~nk~1 by the TPRl gene as follows: plasmid pl7HlD was first produced by incerting the .Sacçh~lolllyces cerevisiase yeast HIS3 gene (1.7 kb BamHI fragmP.nt with ~ ~rtors). (The 1.7kb BamHl fr~gmPnt co..l~;n;.-g HIS3 was ;.CQ1~ted from pYACneo (Traver et al, PNAS 86:5898). pYRP17 is from New Pngl~nd Biolabs) into the XbaI site in the TRPl gene of pl~cmid YRP17. A 3.1 kb ScaI-StuI 5 f ~,~ from pl7HlD co,~ ;n;~g pBR amp s~P~uPnC-p~s~ all TRPl sequ~-nces 5' of the Xba site, the entire HIS3 gene, and 657 bp of TRPl se~uPn~s 3' of the XbaI
site, but not in~ li~ an intact ARS el( -"~"1 was gel pllrifiP1 YAC y2~ shares homology to all s~lue~-r~s in this purified r.,.g~.f nl, except for the HIS3 se~uP-n~P,s.
The yeast strain co..l;..n;ng y2-4 was then lla~r~,l.ned with the ScaI-StuI f~gm~P.nt, 10 and subse~uP-ntly HIS+ tran~Ço~ s were SPlP~ctP~d and then scl~ned for the tIp-phenotype and for linkage (by mitotic loss assays) of the HIS+ and URA+
ph~n~lylJes. Three gP.nP,tic~lly correct lla-~sr~ were identifiPl1 (y3-9, yA-l and yA-2) and s~l to the same Southern blot analyses as des~.il~l above for YAC y2-4, which c~.-r;.. -.P~ that these HIS+ YAC d~ ali~es had the desired 15 U~lU1~. Mitotic stability of YAC illt~lily was d~Pmo~-cl-, lPA by growth of the YAC strain under s~l~;ol. followed by structural analysis. YAC l~ n;nn (dele-...;l-~d by mitotic loss assays) was a~A;.--~-Ply 50% for the YNN~-HIS
YAC and 70~ for the Jl.3Lys YAC.
d. Col.~u~ n of YAC NS10-B14. Figure 2 provides a par~al 20 ~s~ n map of YAC NSl~B14. YAC NS10-B14 co.~ ;ns as its insert the same 10.5kb NdeI-SpeI 3' le~ ql fr~gmP.nt of Jl.3LYS (i.e. the C~ con~nl region s~Pgmpnt) ~joinP~ to an apl,lo,~ ely 14 kb Bam~II fr~rnPnt derived from the 5' end of the insert of Pl-570 YAC. Pl-570 co~ ins Pl-570 is a Pl clone from ~Jen~me Systems, identified in a screen for human C y seq~l~Pn~ ~s. To pl~ an 25 input YAC having on its arms splp~hhlp ...~.L~.~ that can be used when in~ln~ n of a genetic selection step is desired, and in particular one that is co...p~;l.lP. in co...l~ n with YAC Jl.3LYS, the TRPl "right" arm (in this case the c~ ulll~.;c aIm) marker of YAC NS10-B14 was mo~ifiP~l to become HIS3+ (and col~u~-~11y trpl-) by gene disl~~lioll as ~lrul,lled in PY~mrle lc. Mitotic 30 stability of YAC il-t~li~y was tlPmo~ t~ by growth of the YAC st~in under 6.
selection followed by structural analysis. YAC l~ lenlion (de~ ~ by mitotic loss assays) was ap~ A~ply 85 % for the NS10-B14 YAC.
e. Constluction of YAC Pl-570-2-1. Figure 3 provides a par~al restri~ti--n map of YAC Pl-570-2-1. YAC Pl-570-2-1 was derived from the Pl clone Pl-570 as follows. The 85kb SalI-NotI insert fr~gm~nt from Pl-570 was 5 i~olqtçd and cloned as a YAC using the BamHI-NotI centric alm and a SalI-BamHI
f~gm~.nt of the ~~Rntrir arm. During the yeast L~ sru...~;r,n, a de1~ti-n of about 26kb from the C yl region of the insert had oc~;ull~l, resl-lting in a YAC of about 70kb. The 5' end of the insert (co.l1~ini~ the 14kb Bam~ fragment of NS10-B14) was oriented ~dj~e.nt to the ~~~ntric YAC alm. This YAC, Pl-570-2-1, which 10 co~ ls the C y3 region, is then m~ifi~l to contain a mouse 3' enhqnrer obt~inPd as a 15 kb r.Ag...-.~ll from a mouse gem~mir phage library (Clontech).
f. Construction of YAC C13-X15. Figure 4 provides a pafial rçst irtion map of YAC C13-X15. YAC C13-X15 co ,l~ins as the 3' (and u~,e~ )p;i~) portion of its insert the a~l,lu~ p~ly l5kb region of Jl.3LYS that 15 co..1~;nc the VH6 region (Choi et al., Nat. Genet. 4, 117 (1993) which is herebyillcol~ldted by l~fe~Ace) adjoined to a 13kb t~ f ~ll from YAC13.3.
YAC13.3 was i~ol~ed from the ICRF YAC lib~ary (Larin et al., Proc. Natl. Acad.
Sci, 88:4123-4127 (1991)) using a mixed probe for human ;~ u)~ll)blllin heavy chain VH3 s~.,~ ,s as probe. The miYed probe was produced by PCR using the 20 following oligos as spe~ified for VH3 genes by Campbell et aL, ~ol~ulqr )logy vol 29: 193-203 (1992):
VH3-ldr: 5'-CCATGGA~ ~ l ~GGCTGAGC-3 ' [SEQ ID NO: 5]
VHFR3-COM: 5'-CAGTAATACACGGCC~l~l~ -3' [SEQ ID NO: 6]
To ~ an input YAC having on its arms se1~t-q-hle ...-.1.~...~ that 25 can be used when in~ si~n of a genetic s~ ti- n step is desired, and in particular one that is cG...p~il,le in comhinqti-n with YAC Jl.3LYS, the TRPl "right" arm (in this case the ~nl,.,~ .;c arm) marker of YAC C13-X15 was modified to beco...e HIS3+ (and c~ ently trpl~) by gene disruption as p~.r~ ed in FYqmpl~ lc. Mitotic stability of YAC illt~lil~ was demo~ ed by growth of the 30 YAC s~lrain under selection followed by stTuctural analysis. YAC retenti~n (dete.. ;~ l by _itotic loss assays) was a~ P~ly 100% ~r the C13-X15 CA 02204359 l997-05-02 37.
YAC, sugge~ing that multiple copies of the C13-X15 YAC were present in each h~r1Oit1 yeast cell.

pl~ 2. Meiotic Reco",billalion Between YAC Jl.3Lys and YAC YNN~y-HIS.

a. Figure 1 provides a sçh~ de~ g the recomhinqtinn event be~ween YAC Jl.3Lys and YAC YNN~-HIS and the G~l~t~ l~cc...l i~nl YAC products.
The desired YAC in this particular case is the larger recombinqnt product YAC
(ap~ y 125 kb) d~P~iErqtP~l YAC Jl.3-ye2. By this recomb;~. ~iOI~ event heavy COll5~ chain yl s~luç~t,es have been added to the cloned SpeI-SpeI of 10 u~.... ,-.-ged human i~ .u~ylobulin heavy chain locus i.. ~lnbulin ~H) chain gene to create a more cc ~ e heavy chain mini-locus. As will be d~Pmo~.0-. 1~ inthe following G~;~lllplFs, the .n~ll.t~s of the present invention enqhlP,1 a~t1itionql regions of the gene locus to be q.~P.mh1Pd in the order found nqhlr.qlly oc~ on the chlr....~sso...e. In ~ldition~ as will be seen, genomic regions were ju~l~posed (in lS
a ru-.tl;on~ r) despite the lack of a YAC clone con~ g a particular intervening region of overlap by l~",l~lion using gPnPtit~lly e~ Fe-~d "lin~ng" YACs.
b. Mating and Sporulation ~locol. Haploid parents were grown under co.-~;l;. n~ se1~cti~ for and ...~ ing the input YAC. For the mass mating, cells 20 (--5x107) from each parent culture were mixed, pell~ted, washed, and s~
was dera.-1ed Cells were res~ls~ d in a ...;n;...~l volume of liquid YPD (or more simply the residual water after der~ ) and the cell slurry was spotted to aYPD plate (non s~l~1ive for the eYpected diploid). Tnruk~ion was pe.rul",ed at 30~C for 4-5 hrs, after which cells were spread across the entire plate and ;~ "b~l~A 25 at room 1~1ll~ alul~ 20-25~c for about 18-24 hrs. Cells were Se~APed into sterile water and c~--..1t~A About 108 cells were po.l11o.tÇd r~s~ n.1eA in 10 mls of liquid sporulation m~Aillm, and il~ b~leA at 30~C for 4-5 days with occasional momlul~gfor etfi~ient sporulation. Sporulation terhniques (incl~ in~ sporulation, spore P~ l.. l,.,1 sporei~ lion,etc)areall~e~.;~ inGuthrieandFind,eds. 30 Me~th~ in Enzymology v194:94-109, 146-149. Spores were i~ol~ted (l?ocl~mill et al. (1991) in l~eth~s in Enzymology 194:147-149 (which is hereby incol~olaled by WO 96/14436 rCT/US95/14966 - 38.
reference) and the spore-enrirh-p~l population was plated to media selective for the desired recombinant YAC (for ~ k~ LYS2 and ~S3 on the arms of the desired YAC in this G~ ) and cou,ll~r-scle~tive for the undesired recombin-qnt YAC and one of the parental YACs, YNN~y HIS (against URA3 in this eY~mple). Cs~lonip~s arising on the selective media were scl~ned by genetic and by physical methods for S
the p~s~nce of the desired recombinqnt YAC.
c. Scl~n.ng and Ch-q-r~çt~ YACs. Since the mol,c~llqr chqr~q~ ;r n of the cqn(~ qtp YAC clones require p~e~"-lir)n of agarose co.~li il-in~ large molP~Illqr weight DNA blocks of each clone (DT Burke, GF Carle and MV Olsen SciPnce 244:1348 (1987)), a high-throughput procedure for 10 g~n.~ n of small mlmhP~ of blocks from mllltir1P strains was developed.
Briefly, ap~ p~ly 10 mls of each culture was grown to late log or early s~ phase, and the cells were pelleted at lOOOg for 10 ...;n-~les. After ~ec~ g off the me linm~ the pellets were le~ Pn-led in the residual mPAinm (appluA;~ tely 109 cells in 300~1). Ap~u~ .ly 150~1 of a 2% low melting point 15 agarose sol~ltinn in 50_M EDTA was quickly added, and a single large well of theBio-Rad well molds were filled with the IlliAlUl~ (Bio-Rad Lab~ t-,. ;es, Hercules, CA) The blocks were transferred to a 6-well tissue culture plate, each well conl~;n;n~ S mls of a 4mg/ml solntinn of lluvo;c~ e 234 (Novo Nordisk) in lM
Sorbitol, 50mM KPi, 100mM EDTA, pH 5.5. After a 60 minute ilu ~ at 20 37 C, the novozyme soh~lion was removed by ~ ioll and the blocks were washed in 100mM EDTA, 1 % Li-dodec~l-sulfate, 1 % ~Lusyl at 37 C for 40 .~.;.--,les. The LiDS-sa~ l wash l~ul~l twice using fresh LiDS~ l each time. The blocks were then washed once in 50mM EDTA, and loaded onto pulsed field gels. After CHEF PFGE (Bio-Rad), the gels were EtBr stained with ethi~ lm 25 bromide, UV nicked, NaOH denalul~d, and transferred in 0.5M NaOH l.5M NaCl for 2 hours by capill~y transfer onto Hybond N+ posilively ch~g~ nylon m~-..h" ,~. The blots were hybridi_ed from 1 hour to OVG~ ght in hybri~i7~tion solution co ~1Ai..;ng 10% dçYtr~n sulfate and 1~11~ probe. After hybri~
blots were washed, and GA~)OSed to autoradio~;~a~hic film from 10 ...i.- ~les to 1 30 hour. Using this stre~mlin~d ylulucol, one person could screen 100-200 c~n~ te YAC clones at a time over a period of two or three days. This rate is at least a 10-39.
fold inc~q~P, in sample throughput over previously des~ e~ procedures. Results of the recombination are p,~3~nled in Table 1 below.

PYqm~le 3. Meiotic Recombinq-tinn bel~n YAC Jl.3Lys and YAC NS10-B14.
Figure 2 provides a scl~ depi~ the recomhinqtion event l~l~cen YAC 5 Jl.3Lys and YAC NS10-B14 and the e~ ;led recombinqntYAC products. The desired YAC in this particular case is the larger recombin. nt product YAC
(a~ v~ Ply 122 kb) de-sigl~q~-Ptl YAC Jl.3-B14. The mating, sporulating, and genetic selP~tion p~ocedul~s were as des~.il.~l in PYqmr1P 2. The dirr~ ce b~lween the recomlJi.. ~;on~ desc.;l~ed in PY~ plf 2 and PYqmrlP 3 is that the 10 region being added (ie, the dowlls~ end of the smaller YAC) is dirr~ . In PYqmrlPS 2 and 3, the u~ wll YACis Jl.3LYS. In ~.;....plPs 2 and 3, the region of overlap is the lO.Skb N-S fr.q.~nPnt In ~Y;...~l,lr 2, the desir_d ~collll)i~nl YAC has the 17kb C yl region of p~e2 added to the 3' end of Jl.3LYS, wL~"~,as in~,~...pl~ 3, the desired recomhinq-nt has the 14kb BamHI f~gmP.nt derived from Pl- 15 570 added to the 3' end of Jl.3LYS. Results of the l~o...h;~- ~;on are p,~ ed inTable 1 below.

l~ullple 4. ~eitoic Recolllbil~d1ion l~lweenYAC Jl.3-B14 and YAC Pl-570-2-1.
Figure 3 provides a srh ~ ;c d 1;ng the recomhinqtion event bel~ow~ YAC 20 Jl.3-B14 clone #3 and YAC Pl-570-2-1 and the ~ cted l~col~i"anl YAC
products. The desired YAC in this p~rtirll1qr case is the larger recombin. nt product YAC de~i$l~"PA YAC Jl.3-570. The mating, spolulating and genetic sPl~ti~n steps were pelroAlllGd as des_,;l)Gd in Example 2, except that selection con-liti~n~
for the Lys 2 and Trp 1 se1Prtq-hle ".a,1~ on the desired recombinant YAC were 25 osed (rather than for Lys 2 and His 3). Se~ n against Ura 3 was also p~ro~llAGd. Results of the recombin~ n are p,csG,l~ed in Table 1 below.

~nple 5. Meiotic Recomhinqtion l~w~n YAC C13-X15 and YAC Jl.3~ye2.
Figure 4 provides a s~ lir d~;e1;~g the recomhinqti~n event l~lwe~l~ YAC C13-X15 and W O 96/14436 PCTrUS95/14966 40.
YAC Jl.3-ye2, and the G~l ected recombinant YAC products. The mating, spomlation and se1~ti~ n were pelrolllled as desr~ ed in Example 2, except that se1ecti- n was for URA3 and HIS3 and against LYS2. R~combin~tion frequency ,s~.ltad in Table l) over the lSkb region of h.. o1~-~y was si~ifi~nt1y lower(a~p~x;~ P,1y 500-fold) than e~pe;t~d, and ap~ 'e1y 20-fold lower t_an the 5 ~co...b;~ n frequen~;es seen in ~Xpe ;~ nt~ given in PY~mp1~s 2, 3, and 4.

~.Y~mrle 6. Meiotic Recoml)it~alion l~n YAC Cl3-XlS and YAC Jl.3-Bl4.
Figure S p~ovides a 5~hr.~ epi-~1;ng the recomhin~tion event l~tween YAC
Cl3-XlS and 10 YAC Jl.3-Bl4, and the G'l~i~e~l reccmb;r.-~;. n products. The mating, spolulation and s~c.l;ol- were pe.rol,l.Gd as ~lesc-rihed in ~.;....l~1e 5.

~y~mFle 7. Su~ of Recollll)inàlion E~e. ;---I -nl~. In Table l meiotic recomhin~tion rl~u~ n~:Fs for 4 crosses involving YACs ~llo~ g diploid 15 inco...lAI;1);1ity are yl~se ~1ed.

41.

RECOMBINATION RATES
C~ ~ S~o~ Pl~t-d f Cobnfe~ t Comct % Adju~d % lAo~h o~ ~b/clU
on ~tion u ~ r ~ . - r . .. ~o~b~, 1. 11.3Ly~ ~ YNN~-HIS 5.1 s 10~ 45 9 0.053 0.17 10.5 63 5 2. Jl.3Ly~ ~c NS10-B14 18 ~t 10~ 85 15 0.025 0.041i 10.5 227 3. Jl.3-B14 ~ 570-2-1 3.7 ~ 10~ 25 9 0.072 0.099 14.5 141 4. Jl.3~c2 ~ C13-X15 1.2 ~c 10~ 30 14 0.0035 0.010 15 1500 The total number of correct rewmhinqnt~ entifif~ as such by showing 10 correct behavior of genetic ~ f . ~, correct size on a pulsed-fiield gel ele~upholesis and co.~ ~;n;~g e~ ed s~ll,ences by Southern blotting) was divided by the total nllmber of spores phted and then multiplied by 3 (the d~f ~ge n-----~r of viable spores per tetrad) to give the " % recom~ .-". An " ~djllct~P~
% recombin~lion" was then c~lclllqtPA as ~l~Pt~ilP~l below. The average meiotic 15 recombinalion frequency for yeast genomic DNA has been reported as a~,u,~;...-t~,ly 3 kb/cM. S~ 1y however, meiotic rccQml);~ over short homologies l~ YACs with short regions of homology was found to be 20-500 fold lower (See Table 1) than that reported for l~co..~l.;n~ n I~lwee~ larger regions of homology. Despite this ~ edly low recomb~lion rlc lu~,~;y, the 20".e~ of the present i"~enlion provide the means to ov~ e this biological as d~ on~ d herein to readily obtain desired recombin~nt YACs by homologous l~co..~bin~ n over l~ ely short homology regions. Factors such as pru~ ily of ce~ u~ s and telomeres, and sp~ific s~l~en~e e1e~ such as reçoml.;n~ n h~)t~t~ may affect lccûl,lbiL~tion frequency. 25 YAC retention (de~ in~l by mitûtic loss assays) was &p~,u,~ oly 50%
for the ~ YAC, 85% for the NS10-B14 YAC and 70% for the Jl.3Lys YAC.
When each h~rloitl culture was el;t"~ on a pulsed-field gel ele~11u~ho~esis, it app~d that ap~lu~ lely 100% of the ~y and NSl~B14 YACs were "intact"
(i.e., of the correçt size); wh~"~s ap~ ;"~ .ly 90% of the J1.3Lys YAC 30 was"intact" by the same çrit~ . Thel~rûl~ of the diploids folmed dunng the 42.
mass mating in cross 1, only 32%, on average, had both intact YACs; while in cross 2 this mlmb~r increases to 54%. These figures were used to derive the "adjusted % recomb;n~ " in the table above. The length of u~ y~ g homologous s~u~nr-Gs b~lw~l the two YACs was 10.5 kb in crosses 1 and 2 and 14 kb in cross 3. The data given in Table 1 is for spores purified from the mass- 5 mated cultures by the mPthod of Rr,rl~mill et al., (1991) in Guthrie and Fink (eds.) Methods in Enymology 194:147-149. Spore pnrifir~tion was close to 100% in both crosses.
In sharp contrast to the results provided by the m~th~s of the il~ iOll desr~ ed herein, in recomhin~tinn ~pe~;...~nt~ ~Iween YNN~y-~S and Jl.3LYS 10 (clone 13D) wh~ l diploid cells were selected (over haploids) by plating to selective media followed by plating to media (ura dlu~ul, trp dlu~ûul) selectivefor each YAC prior to sporulation, wLc~.n ~lip'oids were t;AIel~ively grown, it was found that spores ~UlViVil~g plating on media sele~live for the desired recombinant (his dlupuul, lys dl~l~uul, and FOA con~;ni~-g did not contain the 15 desired l~...h;n~ It was sul~ ly found that the ..~ilo~ lly-grown diploid culture had a~p~nlly lost co...pletely the YNN~y-HIS YAC before ulld~;oing m~;~ si~ despite the sPlectinn con-liti~m~. Accûldillgly, the methods of the i~lv~ ion provide means to produce and isolate desired recombinant YACs despite diploid inco...p~il.;lity ~eell input YACs and despite relatively small regions of 20 homology.

)le 8: Analysis of large fraPm~nts of the human ;.. ~-no~lo~ulin heavy chain gene cloned in YACs. Two YACs cc.n~ g large fra~mPnt~ of the human ;... -- globin heavy chain variable region were lir~n~etl from the ~ l 25 Research Council, IJK (see figure 7, ~t~rh~l). These YACs (T24a.1 and T10.1) together l~p~S~.~ over 50% of the variable gene se~ As depicted in figure 8, T24a.1 is a 460kb YAC co..l~;nin,P an intact genomic fr~ nt enco...~ in,P at least VH3-26 ~uugh VH3-57, and T10.1 is a 370kb YAC which spans at least VH3-15 ~uugl~ VH3-57, but may ca~Ty an ul~r-h-l~ctçri7e~ del.,lion of 30 ayl~lu~ately 150kb. Southern blotting with a c~ ulllenc arm ~ifi~ probe (the 2.7kb BamHI-Pvu~ f~gm.ont of pBR322) indicated a 17kb and a 13kb XhoI

WO 96/14436 PCTIUS95tl4966 43.
te....ilul fr~mPnt for T24a.1 and T10.1 l~e~ively. Since the centric YAC arm conlains a b~ct~Pri~1 origin and a ~ Pm~P, gene, the ~..,..i,.~l portion of the YAC insert ndjlG~-nt to the centric arm can be cloned by vector arm circu1~ri7~ti- n (Nelson and ~U.. IISttiil~, eds. YAC T ih~TiP~s~ A Users Guide, WH I7~ and Co., pp 3-4, (1994)). These fr~Pnt~ were cloned by li~tinn of a XhoI digest 5 of total yeast DNA, electl~polaLion of XLl/blue cells (St-~t~g~nP), and s~lo~l;on on LB+lOO,ug/ml plates. The restrirti- n maps of endclones 24.13 and 10.33 from YACs T24a.1 and T10.1 ~ ely are given in figure 9. These fr~nP.nt~ will be used to constNct linldng YACs which will bridge T24a.1 and T10.1 with Jl.3-~3 and Jl.3-B14. 10 plc 9: ConstNction of T24-X15 linl~nP YAC. The 12kb XhoI-EcoRI
r~ ".r.~1 from 24.13 is j~ol~P~ and su-hclon~d into the SalI-EcoRI sites of pGP2b to create pGT10. The l5kb XhoI fr~nPnt from pJlXK.31 is isol~l and cloned into the XhoI site of pGT10 to create pGT10-X15. A clone co.~ .;n;~g the 5' end lS
of the l5kb XhoI fr~gmP.nt a~ ent to the 12kb T24 fragmPnt is id~ ;r;ecl by restri~ti-m lllayyillg. The 27kb NotI r.~,....~1 from pGT10-X15 is icolqt~ and ligated to pYNN YAC vector arms and llallar~ ed into a yeast host strain such asYPH857 (Yeast t'enPtir-s Stock Center, nf-~ y CA) or AB1380 (W-~;ngl-l-University). Those YACs cQn~ini~g the T24 f~gm~nt ndja~nt to the acentri~ 20 YAC arm can be i~ nl;r,~ by l~s1~ n digest Sou~ analysis, ~les;g~ 4 T24-X15 linking YAC, and recomhinP~ with Jl.3-~y3 and Jl.3-B14 acco~ g to the invention to yield YAC T24-Jl.3~y3 and YAC T24-Jl.3-B14. Recombinants can be sf1~led for URA3 and HIS3 and against LYS2 on media lacking uracil and hi~ti-linP and CC~.~t~;n;ng 0.2% alpha-amino ~- lip~t~. (Chatoo et al., (~enPti~s 93:51 25 (1979))-Example 10: Recombination of T24a.1to YAC T24-Jl.3~y3 and YAC
T24-Jl .3-B14.
LYS2 derivatives of T24a.1 and T10.1 can be ob~i~l~d by Llallsrolllla~ion of 30 T24a.1 and T10.1 yeast host strains with the lOkb ~indm r.,.~ l from pRVl, followed by se~ l- for lysine plo~u~luyhy. Clones c~ ;n;ng the LYS2 pRVl 44.
f~gmPnt targeted into the ~cPntric arm of the T24a.1 and TlO.lYACs can be i-iPntifi~P~l by PFGE Southern blotting using a LYS2 specific probe. The res~lt~nt YAC, T24L l~o...hil~&~ to YAC T24-Jl.3~3 and YAC T24-J1.3-B14 by s-PlP~tin~ for LYS2 and ~S3 and against URA3 on media lacking lysine and hi~tirlin~. and C~n~;n;i-g FOA. YAC T10.1 is .~imil~rly recombined to YAC T10- 5J13S y3 and YAC T10-J13-B14.

Example 11: Constlu~ion of linl~n~ YACs by LR-PCR.
In some cases, the te-...in~l fr~gmPnt~ of YACs nP~Psc~.~ to construct linking YACs will not be readily clonable by vector circ~ ri7~tion. A number of 10 ive mPth~ls have been desc~;l~l (s ~ d in Nelson and r~low~ eill, eds. YACTih~riP-s,A Users Guide, WH Fl~ and Co., pp 102-107, (1994)) but these approaches produce clones of generally less than 3 kb. Although using the methods of the present il~velltion yeast l~...bi~ )n is pos~il,lc over such short tracts, it would be more pl~f~l~d to have o~t;lldp~ing regions of about 15 10kb. Further, the ~c~P.mhly Of ~.Ill;nAl frAgm~.nt~ into a linking YAC ~3Uil~,Sseveral li~tinn and Ll~l~r~ n steps in E. coli and in yeast. In other cases, convenient ~i" - ;c~ sites may not be available for the i~o1~tion of the le~ -..;i~l f~gm,ont from the vector arm portion of the vector circ~ ri7~tion pl~mi~l For eY~nr1e, there are mllltirlP. EcoRI sites within the 8.5kb te ---;n~1 fr~gm~nt of the 20 T10.1 çn~c~ & 10.33, pr~,lntling i~o1~til~n of the entire insert as a XhoI-EcoRI
fr~gm~nt The te-...;.~Al fr~gmP.nt~ of 10.33 and 24.13 were i~ol~tPd by long range PCR (LR-PCR; Perkin Elmer Col~l~Lion, Norwark, CT) using the XL-PCR kit (Perkin Elmer). The following oligos, ~ifi~- for YAC centric 2S
vector arm sequpnces~ were used to amplify the inserts:
LRVT-lF: 5'-CTC TCG AGG GCT TGG TTA TGC CGG TAC T-3' [SEQ ID
NO: 7] and LRVT-lR: 5'-CTC TCG AGC CTC TGA CTT GAG CGT CGA T-3' [SEQ ID
NO: 8]. 30 Bands of about 8.5kb and 12kb were ~...pl;r.P~ from 10.33 and 24.13 ~speclively,and i~olqtPcl as XhoI r.~g...~ of about 8.5kb and 12kb for subseq~l~Pnt cloning steps.

WO 96tl4436 PCT/US95/14966 45.
The linking YAC can be partially constructed by LR-PCR. As depicted in Figure 9, a set of 4 oligos were synth~i7ed and used to amplify the 10.33 and 24.13 insert fr~gmP.nt~ and the JlXK.31 insert fr~gment The primer se~u~-n~es were:
SlAR: 5'-AAG CGG CCG CAT GAA TTC TAT CTG GGA AGT GAA TGG 5 AGA C-3' ~SEQID NO: l];
S3'CX: 5'-TAG CGG CCG CAT TAG AAT TCA GCT GCA TGT GTC AGA
GGT T-3' [SEQ ID NO: 2];
S2'AXl: 5'-ATT ACC CTG TTA TCC CTA GGC CGA ACA GGC AGA CAT
CTG TGA-3' [SEQID NO: 3]; 10 S2CRl: 5'-GGC CTA GGG ATA ACA GGG TAA TAC TCT CGG TAG CCA
AGT TGG-3' [SEQ IS NO: 4].
No~d sites were inrh~ded in Sl and S3', and i-SceI sites were inclllde l in S2 and S2'. The JlXK.31 derived amplific?tion product is joined to the 10.33 or the 24.13 amp1ific~tion products by t1igestion of both products with i-SceI, ligated to 15 form circles, and digested with NotI to form l;n~ d, joined r-,.g-~ . These No~ fr~gment~ are cloned in pYNN vector arms, l,allsro.~,led into yeast, and analyzed for structure by l~ s~ n digest S~u~ analysis.
All publications and patent app~ tinn~ mpntion~d in this s~;r.-~;on are herein incol~ldted by r~ nce to the same extent as if each 20individual publir~tion or patent applic~lion was srecific~1ly and individually in~ qte~ to be inco.~o.db~l by ,~f~ nce. The .~f~l~ces ~ cus~e~ herein are p~ovided solely for their ~1i.~lnsure prior to the filing date of the present appli~tinn. Nothing herein is to be construed as an ~lmi~ m that the inventors are not enthled to ~-~1e~ e such ~iSclos~lre by virtue of prior i~ iom 25 The h~ ti~n now being fully desc.il)ed, it will be dplJal~,lt to one of ol.lin~ ~ sl~ll in the art that m. ny Ch~llgCS and mo lifi~qtions can be made- thereto without dc;pal~ing from the spirit or scape of the ap~ended chims.

W O96/14436 PCTtUS95tl4966 46.
SkyuKNC~ LISTING

(1) GRNRT~T lNrO~_'TION:
(i) APPLICANT: Barbara Dunn Theo~nre K. Choi 5 (ii) TITLE OF lNvhNllON: METHODS FOR M~KING
REC~MRIN~NT YEAST
ARTIFICIAL r~M~SOMES

(iii) N~JMRRR OF ShYUhN~S: 8 10 (iv) COPT~R~ONvKN~b ~nDPRqS:
(A) ~n~pRcSRR COOLEY, GODWARD, ET AL.
(B) STREET: 5 Palo Alto Square, Fourth Floor (C) CITY: Palo Alto lS
(D) STAT_: California (E) COUhl~Y: U.S.A.
(F) ZIP: 94044 (V) Cu ~Ul~K READABLE FORM:
(A) MEDIUM TYPE: 3.5~ Di~kette, 1.44 Mb 20 (B) C~ ~ul~: IBM PS/2 Model 50Z or 55SX
(C) OPERATING SYST_M: MS-DOS (Version 5.0) (D) SOFTWARE: WordPerfect (Ver~ion 5.1) (Vi ) ~UKK~h L APPLICATION DATA:
(A) APPLICATION NCTMRRR 25 (B) FILING DATE: Nov~ 'er 4, 1994 (C) CLASSIFICATION:
(vii) PRIOR APPLICATION DATA:
(A) APPLICATION NUMBER: 07/900,972 (B) FILING DATE: June 18, 1992 30 (C) APPICATION N~MRRR: 08/148,177 (D) FIL_D: N6~ ' ~r 5, 1993 (viii) Al-lOkNhY/AGENT lNhC~ TION:
(A) NAME: Timothy E. Torchia (B) REGISTRATION NnMRRR 36,700 35 (C) K~_N~/~O~Ahl~ NUMBER: GENP-003/OOUS

(ix) TELEC~--r--lNlCATION lNhu~L!TION:
(A) TELEPHONE: (415) 843-5481 (B) TELEFAX: t415) 857-0663 40 (2) lN~O~L!TION FOR S_yu_~ -lhlCATION ~UMRRR: 1:

(i) S_YU_N~_ CHARACTERISTICS: (A) LENGTH: 40 45 W O96/14436 PCTnUS95114966 47.
(B) TYPE: nucleic acid (C) STRPNnRnNRss ~ingle (D) TOPOLOGY: linear (xi) ~hQuhN~h DESCRIPTION: SEQ ID NO: l:
AAGCGGCCGC ATGAATTCTA l~ AAGT GAATGGAGAC 5 (2) lNrv~_!TION FOR ~hyuhN.~r lvh~llrlCATION NUMRRR 2:
(i) ~hyuhN~h CHARACTERISTICS: (A) LENGTH: 40 (B) TYPE: nucleic acid (C) STR~NnRnNRss single l0 (D) TOPOLOGY: linear (xi) ~hyuhN~r DESCRIPTION: SEQ ID NO: 2:
TAGCGGCCGC ATTAGAATTC AGCTGCATGT GTCAGAGGTT

(2) lNrO.~L!TION FOR Shyurl._h lvhNllr-lCATION NUMBER: 3: l5 (i) ~hyur;N~h CHARACTERISTICS: (A) LENGTH: 42 (B) TYPE: nucleic acid (C) STR~NnRnNRSS ~ingle (D) TOPOLOGY: linear (xi) Sr;yuhN~r DESCRIPTION: SEQ ID NO: 3: 20 (2) lNr~ _ TION FOR SKyuhN[~: lvh~ rlCATION NUMBER: 4:
(i) ~hyur~uh CHARACTERISTICS: (A) LENGTH: 42 (B) TYPE: nucleic acid 25 (C) STRANI~K-~NK~S ~ingle (D) TOPOLOGY: linear (xi) Shyuh~._h DESCRIPTION: SEQ ID NO: 4:
GGCCTAGGGA TAACAGGGTA ATAw ~lCGG TAGCCAAGTT GG 42 (2) lNrO.-~_.TION FOR ShOuhN~r. lvr;NllrlCATION N~MRRR: 5:
(i) ~r;yuhNuh CHARACTERISTICS. (A) LENGTH: 20 (B) TYPE: nucleic acid (C) STRPNI-Kl-NK~s ~ingle (D) TOPOLOGY: linear 35 (xi) Sr;yuhN~h DBSCRIPTION: SEQ ID NO: 5:
CCATGGAGTT l~GG~-l~AGC 20 (2) lNrO.-~_.TION FOR ~r,yuhN~h lvhNllrlCATION Nr~MRR~ 6:
(i) Sr;yur;Nur CHARACTERISTICS: (A) LENGTH: 20 (B) TYPE: nucleic acid 40 (C) STRPNnRnNR.~s ~ingle (D) TOPOLOGY: linear (xi) ~r;yuhN~r; DESCRIPTION: SEQ ID NO: 6:
CAGTAATACA CGGCC~-~lC
(2) lNrO~L.TION FOR ~hyurN~r; lvhNllrlCATION NUMBER: 7: 45 W O96/14436 PCTrUS95/14966 48.
(i) ShQUhN~ CHARACTERISTICS:(A) LENGTH: 28 (B) TYPE: nucleic acid (C) STR~NnRnNRSS Bingle (D) TOPOLOGY: linear 5 (Xi ) S~QU~N~ DESCRIPTION: SEQ ID NO: 7:
.clC~AGGG ~l~l-lATG CCGGTACT
(2) lN~O~_!TION FOR SKQU~:N~ ~hNll~lCATION NUMRRR: 8:
Qu~N~h CHARACTERISTICS: (A) LENGTH: 28 (B) TYPE: nucleic acid 10 (C) STR~NnRnNRSS: Bingle (D) TOPOLOGY: linear (xi) Sb:OuhN~ DESCRIPTION: SEQ ID NO: 8:
c~AGCC TCTGACTTGA GC~lCGAT

Claims (21)

49.
WHAT IS CLAIMED IS:

1. A method of producing a recombinant YAC, comprising the steps of (a) mating a first haploid yeast cell comprising a first YAC to a second haploid yeast cell comprising a second YAC having a homology region with the first YAC, to obtain a diploid yeast cell, wherein mitotic doubling of the diploid is less than or equal to 8 doublings, (b) sporulating the diploid and/or its mitotic progeny, to obtain spores, and then (c) identifying spores that comprise the recombinant YAC, or alternatively the steps of (d) mating a first haploid yeast cell comprising a first YAC to a second haploid yeast cell comprising a second YAC having a homology region with the first YAC to obtain a diploid yeast cell, (e) sporulating the diploid and/or its mitotic progeny to obtain spores, and then (f) identifying spores that comprise the recombinant YAC by culturing spores by selecting for growth of spores or cells comprising the recombinant YAC, and optionally, selecting against spores comprising an undesired recombinant YAC and further optionally selecting against spores or cells comprising one or both unrecombined first and second YACs.

2. The method of Claim 1, wherein steps (a) or (d) comprises a mass mating between cultures of the first and second haploid yeast cells.

3. The method of Claim 2, wherein step (c) comprises identifying spores that comprise the recombinant YAC by culturing spores under conditions that select for growth of spores comprising the recombinant YAC, and optionally, selecting against spores comprising the undesired recombinant YAC and further optionally selecting against spores or cells comprising one or both unrecombined first and second YACs.

50.

4. The method of Claim 1, wherein diploid mitotic doubling in step (a) or (d) is less than or equal to 3 doublings.

5. The method of Claim 2, wherein in step (a) or (d) the mating step is less than or equal to 5 diploid-doubling times.

6. The method of Claim 5, wherein the mating step is less than or equal to 3 diploid-doubling times.

7. The method of Claim 1, wherein in steps (a) or (d) mating occurs in the absence of a selection for the diploid comprising both the first and second YACs.

8. The method of Claim 1, wherein in step (f) selecting for growth of spores comprising the recombinant YAC comprises selecting for at least one selectable marker present on the recombinant YAC.

9. The method of Claim 8, wherein selecting for growth of spores comprises selecting for at least two selectable markers that are present on the recombinant YAC but are not present on an undesired recombinant YAC.

10. The method of Claim 1, wherein the homology region between the first and second YAC is less than about 40 kilobases.

11. The method of Claim 10, wherein the homology region is less than about 20 kilobases.

12. The method of Claim 10, wherein the homology region is less than about 5 kilobases.

13. The method of Claim 1, wherein during the mating step cells are maintained at a higher temperature during the initial period of the mating step than during the remaining period.

51.

14. The method of Claim 13, the temperature during the initial period is about 25 to 35°C and the temperature during the remaining period is about 15 to 25°C.

15. The method of Claim 1, further comprising the step of isolating the recombinant YAC.

16. A method of producing a desired recombinant YAC, comprising the steps of (a) obtaining a first haploid yeast cell comprising a first input YAC, (b) obtaining a second haploid yeast cell comprising a second input YAC, (c) obtaining a third haploid yeast cell comprising a linking YAC having a first homology region with the first input YAC and a second homology region with the second input YAC, (d) obtaining an intermediate haploid yeast cell comprising an intermediate recombinant YAC by meiotic homologous recombination between the linking YAC
and the first input YAC, and (e) obtaining a final yeast cell comprising the desired recombinant YAC by meiotic homologous recombination between the intermediate YAC and the second input YAC, wherein steps (a), (b), and (c) are performed in any order.

17. The method of Claim 16, wherein either the first homology region or the second homology region is less than about 40 kilobases.

18. The method of Claim 16, wherein both the first homology region and the second homology region is less than about 40 kilobases.

19. The method of Claim 16, wherein the steps (d) and/or (e) comprise the steps of (1) mating the first haploid yeast cell to the third haploid yeast cell in the case of step (d) to obtain a first diploid yeast cell and mating the intermediate haploid yeast cell to the second haploid yeast cell in the case of step (e) to obtain a 52.

second diploid yeast cell, wherein mitotic doubling of the diploid is less than or equal to 8 doublings, (2) sporulating the first and second diploids and/or their mitotic progeny, to obtain spores, and then (3) identifying spores comprising the intermediate YAC in the case of step (d) and spores comprising the desired recombinant YAC in the case of step (e), or alternatively the steps of (4) mating the first haploid yeast cell to the third haploid yeast cell in the case of step (d) to obtain a first diploid yeast cell and mating the intermediate haploid yeast cell to the second haploid yeast cell in the case of step (e) to obtain a second diploid yeast cell, (5) sporulating the first and second diploids and/or their mitotic progeny, to obtain spores, and then (6) identifying spores comprising the intermediate YAC in the case of step (d) and spores comprising the desired recombinant YAC in the case of step (e) byculturing spores by selecting for growth of spores or cells comprising the intermediate or desired recombinant YAC, and optionally, selecting against spores comprising an undesired recombinant YAC and further optionally selecting againstspores or cells comprising one or both unrecombined first input and linking YACsin the case of step (d) or intermediate and second input YACs in the case of step (e).

20. The method of Claim 16, wherein the linking YAC is constructed by a method comprising the step of (a) obtaining a first vector having the first homology region and a second vector having the second homology region, and (b) amplifying the first homology region and the second homology region by long-range polymerase chain reaction.

21. The method of Claim 20, wherein the amplification step comprises (a) extending a first primer able to bind to a region upstream of the first homology region, 53.

(b) extending a second primer able to bind to a region downstream of the first homology region, (c) extending a third primer able to bind to a region upstream of the second homology region.
(d) extending a fourth primer able to bind to a region downstream of the second homology region, wherein the first and fourth primers each comprise a first restriction enzyme site sequence that is not present in either amplified homology region but is present as a cloning site in a linking YAC vector, and wherein the second and third primers each comprise a second restriction enzyme site sequence that is different from the first restriction enzyme site sequence and that is not present in either amplified homology region nor in the linking YAC
vector arms.